Skip to main content

Full text of "soil analysis"

See other formats

Replenishing Soil 
Fertility in Africa 

SSSA Special Publication Number 51 

Related Society Publications 

Methods of Soil Analysis: Part I — Physical and Mineralogical Methods 
Methods of Soil Analysis: Part 2 — Microbiological and Biochemical Properties 
Methods of Soil Analysis: Part 3 — Chemical Methods 

1 '/ i , nia, i \ Agn iii n econd Edit n 

Vlin i I i Si I Enviro \iiu in , i i. dition 
Nitrogen in Agricultural Soils 
Nitrogen in Crop Production 
Potassium in Agriculture 

Soil Fertility and Organic Matter as Critical Components of Production Systems 
Soil Testing and Plant Analysis, Third Edition 
Sulfur in Agriculture 

For information on these titles, please contact the ASA, CSSA, SSSA 
Headquarters Office; Attn.: Marketing; 677 South Segoe Road; Madison, WI 
53711-1086. Phone: (608) 273-8080. Fax: (608) 273-2021. 

Replenishing Soil Fertility 
in Africa 

Proceedings of an international symposium cosponsored by Divisions A-6 
(International Agronomy) and S4 (Soil Fertility and Plant Nutrition), and the 
International Center for Research in Agroforestry, held at the 88th Annual 
Meetings of the American Society of Agronomy and the Soil Science Society of 
America, Indianapolis, Indiana, 6 November 1996. 

Roland J. Buresh, Pedro A. Sanchez, and Frank Calhoun 

Organizing Committee 
Pedro A. Sanchez and Roland J. Buresh 

Editor-in-Chief ASA 
Jerry Hatfield 

Edito r- in - Ch iefSSSA 
Jerry M. Bigham 

Managing Editor 
Dave M. Krai 

..Associate Editor 
Marian K. Viney 

SSSA Special Publication Number 51 

Soil Science Society of America 

American Society of Agronomy 

Madison, Wisconsin, USA 



Cover Design: Conrad Mudibo and Darnary Odanga 

1c1?aFTTbr a r y 


Any and all uses beyond the limitations of the "fair use" provision of the 

la i quit ' in 11 i in, ion fi nil til ti lishei i nd/oi the authors); 
not applicable t ntribntions pi I ffict npl til 

U.S. Government as part of their official duties. 

American Societj of Agronomy. Inc. 
Soil Science Society of America. Inc. 
677 South Segoe Road, Madison, WI 53711 

n Number: 97-77363 





Conversion Factors for SI and Non-SI 

1 Soil Fertility Replenishment in Africa: An Investment 
in Natural Resource Capital 

Pedro A. Sanchez, Keith D. Shepherd, Meredith J. Soule, 
Frank M. Place, Roland J. Buresh, Anne-Marie N. Izac, 
A. Uzo Mokwunye, Fred R. Kwesiga, Cyrus G. Ndiritu, 
and Paul L. Woomer 

Soil Fertility in Africa Is at Stake 

Eric M. A. Smaling, Stephen M. Nandwa, 

and Bert H. Janssen .47 

Soil Fertility Management in Africa: A Review 
of Selected Research Trials 

Mateete A. Bekunda, Andre Bationo, and Henry Ssali .63 

A Fertilizer-Based Green Revolution for Africa 
Marco A. Quin'ones, Norman E. Borlaug, 
and Christopher R. Dowswell , 81 

Soil Profile Alteration under Long-Term High-Input Agriculture 

Stanley W. Buol and Michael L. Stokes 97 

Building Soil Phosphorus Capital in Africa 
Roland J. Buresh, Paul C. Smithson, 
and Deborah T. Heliums Ill 

Building Soil Nitrogen Capital in Africa 

Ken E. Giller, Georg Cadisch, Costas Ehaliotis, 

Edward Adams, Webster D. Sakala, 

and Paramu L. Mafongoya . ., 151 

Combined Use of Organic and Inorganic Nutrient Sources 
for Soil Fertility Maintenance and Replenishment 

Cheryl A. Palm, Robert J. K. Myers, 

and Stephen M. Nandwa 193 

Gender and Soil Fertility in Africa 

Christina H. Gladwin, Ken L. Buhr, Abe Goldman 

Clifton Hiebsch, Peter E. Hildebrand, Gerald Kidder, 

Max Langham, Donna Lee, Peter Nkedi-Kizza, 

and Deirdre Williams 219 

Ecological Economics of Investing in Natural 
Resource Capital in Africa 

Anne-Marie N. Izac 237 


The 1996 World Food Summit highlighted sub-Saharan Africa as the 
remaining region in the world with decreasing food production per capita. The 
worst levels of poverty and malnutrition in the world exist in this region. 

A team of scientists has identified declining soil fertility as the fundamen- 
tal agronomic cause for declining food productivity in Africa. A "Soil Fertility 
Initiative for Africa" has been created by a group of international organizations 
including the World Bank, Food and Agriculture Organization (FAO), 
International Center for Research on Agroforestry (ICRAF), International 
Fertilizer Development Center (IFDC), International Fertilizer Association (IFA), 
and International Food Policy Research Institute (IFPRI). 

This SSSA-ASA Special Publication is a comprehensive treatise that brings 
science to bear on current approaches to replenish soil fertility in Africa. It is the 
proceedings of an international symposium cosponsored by ASA Divisions A-6 
(International Agronomy) and SSSA Division S-4 (Soil Fertility and Plant 
Nutrition), and ICRAF. The symposium was held on 6 November 1996 at the 
joint annual meeting of ASA and SSSA in Indianapolis Indiana. 

This book is an excellent resource for anyone interested in the development 
of food supplies in Africa or problems of soil fertility anywhere in the world. Our 
Societies are proud to publish this pioneering work written by its members and 
by colleagues from other disciplines. It is an example of our Societies' contribu- 
tion to addressing problems on a global scale. 

D. Keith Cassel, President 
Soil Science Society of America 

William McFee, President 
American Society of Agronomy 


Sub-Saharan Africa is the last continent facing massive problems of food 
security because of decreasing per-capita food production. Extreme poverty (per 
capita incomes of <1 U.S. dollar per day), widespread malnutrition and massive 
environmental degradation are direct consequences of a policy environment that 
results in large scale nutrient mining. A series of awareness-raising events during 
the last 2 yr has brought the issue of soil fertility depletion in Africa to a similar 
level of importance as the issue of low-yielding rice and wheat varieties in Asia 
three decades ago, which triggered the Green Revolution that dramatically 
increased per capita food production in Asia and Latin America. 

This SSSA Special Publication brings together the current thinking of a 
multidisciplinary team on soil scientists, agronomists, economists, anthropolo- 
gists, and foresters as well as leaders of African national research institutes, inter- 
national research centers, nongovernmental organizations, and universities in 
Africa, the USA and Europe. The 10 chapters are coauthored by a diverse group 
of 41 scientists from 14 countries, about one-half of the authors are currently 
working in Africa. 

The first chapter presents the new conceptual approach of replenishing soil 
fertility as an investment in natural resource capital. It is followed by an analysis 
of the magnitude of soil fertility depletion (Smaling et al.), a review of field 
research trials (Bekunda et al.), an exciting NGO approach (Quinones et al.), and 
a perspective from temperate -region soils (Buol & Stokes). These in turn are fol- 
lowed by three process-oriented chapters on phosphorus (Buresh et al.), nitrogen 
(Giller et al.), and combining organic and inorganic nutrient inputs (Palm et al.). 
The last two focus on key socieconomic considerations, gender (Gladwin et al.) 
and environmental economics (Izac). 

The editors wish to acknowledge the assistance of P.K. Nair and Tom 
Doerge, A-6 and S-4 Division Program chairs for assistance in organizing the 
symposium, to Paramu Mafongoya and Rob Bertram for chairing the sessions, 
the 200 partipants for lively discussions, the 25 peer reviewers for their critical 
input, and to Helen van Houten, Kellen Kebaara, and Fiona Chandler of ICRAF, 
and Marian Viney of ASA Headquarters staff for editorial assistance. 

Roland J. Buresh and Pedro A. Sanchez 


Nairobi, Kenya 

Frank G. Calhoun 

Ohio State University 

Columbus, Ohio 


Edward Adams 

Andre' Bationo 

Mateete A. Bekunda 

Norman E. Borlaug 

Ken L. Buhr 

Roland J. Buresb 

Christopher R. Dowswell 

Costas Ehaliotis 

Ken E. GiUer 

Research Assistant, Department of Biological Sciences Wye 
College, University of London. Wye, Ashford, Kent TN25 5AH 
England. Phone: 44-1233-812401, Fax: 44-1233-813140, Email: 

Soil Scientist, Internationa] Fertilizer Development Center 
(IFDCyinternational Crops Research Institute for the Semi-Arid 
Tropics (ICRISAT), B.P. 12404, Niamey, Niger. Phone: 227- 
722529, Fax: 227-734329 Email: 

Senior Lecturer, Department of Soil Science, Makerere University 
P.O. Box 7062, Kampala, Uganda. Phone: 256-41-540707, Fax: 
256-41-543382, Email: de. 

President, Sasakawa Africa Association, c/o CIMMYT, Lisboa 27, 
Apdo 6-641, 06600 Mexico D.F., Mexico. Phone: 52-5-9230417, 
Fax: 52-5-9541069, Email: 

Assistant Professor, Agronomy Department, University of Florida, 
Gainesville FL 32611. Phone: 352-392-1823, Fax: 352-392-7248, 

Professor, Department of Soil Science, North Carolina State 
University, Box 7619, Raleigh, NC 27695-7619. Phone: 919-515- 
2388, Fax: 919-515-2167, Email: Stanley 

Principal Soil Scientist, International Centre for Research in 
Agroforestry (ICRAF), P.O. Box 30677, Nairobi, Kenya. Phone: 
254-2-521450, Fax: 254-2-521001, Email: 

Senior Research Fellow. Department of Biological Sciences. Wye 
College, University of London Wye, Ashford, Kent TN25 5AH 
England. Phone: 44-1233-812401, Fax: 44-1233-813140, Email: 

Director of Program Coordination, Sasakawa Africa Association 
c/o CIMMYT, Lisboa 27, Apdo 6-641, 06600 Mexico D.F., 
Mexico. Phone: 52-5-9230417, Fax: 52-5-9541069, Email: 

Postdoctoral Research Fellow, National Agricultural Research 
Foundation, Institute of Kalamata, 85 Lakonikis Str., Kalamata 
24100, Greece. Phone: 30-721-91984, Fax: 30-721-27133, Email: 

Professor, Department of Biological Sciences, Wye College, 
University of London, Wye, Ashford, Kent TN25 5AH England. 
Phone: 44-1233-812401, Fax: 44-1233-813140, Email: k.giller@ 


Christina H. Gladwin 

Abe Goldman 

Deborah T. Heliur 

Clifton Hiebsch 

Peter E. Hildebrand 

le-Marie N. Izac 

Professor, Food and 1 i Economics Dej utm nl f >\ 1 10240. 

Institute ofFood and Agricultural Science;; (niversity of Florida, 
Gainesville FL 32611. Phone: 352-392-1881, Fax: 352-392-9898, 

Associate Professoi ( i In I p inn nl University of Florida, 
Gainesville FL 32611. Phone: 352-392-0494, Fax: 352-392-8855, 


Scientist-Soil Fertility. International Fertilizer Development Center 
(IFDC), P.O. Box 2040, Muscle Shoals, AL 35662. Phone: 205- 
381-6600, Fax: 205-381-7408, Email: 

Professor, Agronomy Department, Box 110500, 
University of Florida, Gainesville FL 32611. Phone: 352-392-6187, 
Fax: 352-392-1840, Email: 

Professor, Food and Resource Economics Department, Box 110240, 
Institute of Food and Agricultural Sciences, (.'niversity of Florida, 
Gainesville FL 32611. Phone: 352-392-5830, ext 436, Fax: 352- 

3')A8<>>4, Email: 

Director of Research, International Centre for Research in 
Agroforestry (ICRAF), P.O. Box 30677, Nairobi, Kenya. Phone: 
254*2-521450, Fax: 254-2-521001, Email: 

Senior Lecturer in Soil Science. Department of Soil Science and 
Plant Nutrition, P.O. Box 8005, Wageningen Agricultural 
University, 6700 EC Wageningen, the Netherlands. Phone: 31-317- 
483842, Fax: 31-317-483766, Email: bert.ianssen@bodvru.benp . 

Gerald Kidder 

Fred R Kwesiga 

Paramu L. Mafongoya 

Fax:352-392-3902 Email kidde 

Coordinator, Southern Africa Regional Programme, International 

• ends i • I earch u V- i ti il' ! \l Makoka Research 

Station, P.O. Box 134, Zomba, Malawi. Phone: 265-534277, Fax: 
265-534283, Email: makoi 

Professor, Food and Resource Economics Department, Box 110240, 
Institute of Food and Agricultural Sciences. I "niversity of Florida, 
Gainesville FL 32611. Phone: 352-486-1632, Fax: 352-392-9898, 


t Professor, Food and Resource Economics Department, 
Box 110240, Institute of Food and Agricultural Sciences, 
University of Florida, Gainesville FL 32611. Phone: 352-392-4991 
ext. 224, Fax: 352-846-0988, Email: 

Soil Scientist, Zaml la-ICI 1 roforesi I .seaich Project, c/o 
Provincial Agriculture Office, Provident House, P.O. Box 510046, 
Chipata, Zambia. Phone: 260-62-21404, Fax: 260-62-21404, Email: 
zamicraf @ 

Director, Institute foi Natural Resources in Africa, United Nation 
University, Private Mail Bag. Kotoka International Airport, Accra, 
Ghana.Phone: 233-21-500396, 233-21-500791, Fax: 233-21- 
500792, Email: UNUINRA@ 


Robert J. K. Myers 

Stephen M. Nandwa 

Cyrus G. Ndiritu 

Principal Scientist. Internationa] Crops Research Institute for the 
Semi-Arid Tropics (ICRISAT), Patancheru P.O., Andhra Pradesh 
502 324, India.Phone: 91-040-596161, Fax: 91-040-241239, Email: 

Head, Soil Fertility and Plant Nutrition Programme, Kenya 
Agricultural Research Institute, National Agricultural Research 
Laboratories, P.O. Box 1473 '■'•. Nairobi. Kenva. Phone: 254-2- 
446722. Fax: 254-2-443376. Email: soiK" 

Peter Nkedi-Kizza 

Professor, Soil and Water Science Department, Box 110290, 
University of Florida, Gainesville FL 32611. Phone: 352-392-1951, 
Fax: 352-392-3902.Email: 

Senior Scientific Officer, Tropical Soil Biology and Fertility 
Programme (TSBF), P.O. Box 30592, Nairobi, Kenya. Phone: 254- 
2-622657. Fax: 254-2-521 159, Email: cheryl. 

Frank M. Place 

Senior Policy Economist. International Centre for Research in 
Agroforestry (ICRAF), P.O. Box 30677, Nairobi, Kenya. Phone: 
254-2-521450, Fax: 254-2-521001, Email: 

Webster D. Sakala 

Pedro A. Sanchez 

Keith D. Shepherd 

Eric M.A. Smaling 

Country Director for Ethiopia ami Eritrea. Sasakawa-Global 2000 
P.O. Box 12771, Addis Ababa, Ethiopia. Phone: 251-1-510584, 
Fax: 251-1-510891, Email: 

Agronomist, Maize Commodity Team. Chitedze Agricultural 
Research Station, Department of Agriculture, P.O. Box 158, 
Lilongwe, Malawi. Phone: 265-767222 

Director General. International Centre foi Research in Agroforestry 
(ICRAF), P.O. Box 30677, Nairobi, Kenya. Phone: 254-2-521003, 
Fax: 254-2-520023, Email: 

Senior Systems Agronomist, International Centre for Research in 
Agroforestry (ICRAF), P.O. Box 30677, Nairobi, Kenya. Phone: 
254-2-521450, Fax: 254-2-521001, Email: 

Head, DLO Development Cooperation Programme, Research 
Institute for Agrobiology and Soil Fertility (AB-DLO), P.O. Box 
14,6700 AA Wageningen, the Netherlands. Phone: 31-317-475762, 
Fax: 31-317-475787. Email: e.m.a s 

Paul C. Smithson 

Laboratory Manager. International Centre for Research in 
Agroforestry (ICRAF), P.O. Box 30677, Nairobi, Kenya. Phone: 

254-2-521450. Fax: 254-2-52 1 60 1, Email: :n)ithson@cg 

Meredith J. Soule 

Economist, International Centre for Research in Agroforestry 
(ICRAF), P.O. Box 30677, Nairobi, Kenya. Phone: 254-2-521450, 
Fax: 254-2-521001. [..mail: e (c' 

Senior Principal Research Officer & Head of Soils Programme, 
Kawanda Agricultural Research Institute, P.O. Box 7065, Kampala, 
Uganda. Phone: 256-41-567158, Fax: 256-41-567635, Email: 


Michael L. Stokes Graduate Student, Department of Soil Science, North Carolin; 

University, Box 7619 WMS, Raleigh, NC 27695-7619. Phone 
515-3203, Fax: 919-515-2617, Email: mike@pop.s ' 

Deirdre Williams M.S. Graduate, c/o Adele 

Place, Yorba Linda, CA 9 

Paul L. Woomer Visiting Lecturer, Department of Soil Science, University of 

Nairobi, P.O. Box 47543, Nairobi, Kenya. Phone: 254-2-520328, 

Ll)].:|l l_ 

into Column 2, 

multiply b\ 

Conversion Factors for SI and non-SI Units 

Column 1 SI Unii 

convert Column 2 
into Column I, 
multiply by 

square millimeter, n 

•, L (10" ! 
', L (IO" 3 
', L(10"'i 
', L (10- 
', L (10" 3 
', L (10" 3 

cubic foot, ft 3 


quart (liquid), qt 

cubic foot, ft 3 


ounce (fluid), oz 

pint (fluid), pt 

2 .20 x 10-' 

gram, g(10 5 kg) 

3.52 x 10- 2 

gram, g (10- ! kg) 

2 205 

kilogram, kg 


kilogram, kg 

1.10 x 10" J 

kilogram, kg 


megagram. Mg (tonne) 

pound, lb 
quintal (metric), q 
ton (2000 lb), ton 

Yield and Rate 


kilogram per hectare, kg ha- 

pound per acre 

, lb acre- 1 

7.77 x 


kilogram per cubic meter, kg m- ! 

pound per bushel, lb bu- 

1.49 \ 


kilogram per hectare, kg ha- 

bushel per aci 

. oo lb 

1.59 x 


kilogram per hectare, kg ha- 1 

bushel per acr 

. 56 lb 

1.86 \ 


kilogram per hectare, kg ha- 1 

bushel per acr 

. 48 lb 


liter per hectare, L ha- 1 

gallon per acre 


tonnes per hectare, t ha- 1 

pound per acre 

, lb acre-1 

so -; 

megagram per hectare, Mg ha- 1 

pound per acre 


megagram per hectare, Mg ha- 1 

ton (2000 lb) per acre, ton a 


meter per second, m s- 1 




square metei per kilogram, nf kg" 1 

square centim 

ter per gram 


square meter per kilogram, in kg- 

square millim 

ter per gram. 



megapascal, MPa (10* Pa) 



megapascal, MPa (10* Pa) 



megagram, per cubic meter, Mg m~ 

gram per cubi 


2.09 x 


pascal, Pa 

pound per squ 

ire foot, lb ft 

1.45 x 


pascal, Pa 

pound per square inch, lb in 



Conversion Factors for SI and non-SI Units 

nto Column 2, 

multiph by 


1.00 (K-273) 

Kelvin, K 

(9/5 °C) + 32 

Celsius. C 

9.52 x lO- 4 

joule, J 


joule, J 


joule, J 


joule, J 

2.387 x 10- 5 

joule per square 

10 s 

newton, N 

1.43 x 10 3 

watt per square 

convert Column 2 
into Column 1, 

multiph b\ 

Vork, Quantity of Heat 

calorie per square 

calorie per square 

Transpiration and Photosynthesis 

nilligram per square meter second, gram per square decim 

milligram (H 2 0) per square meter 
second, mg nr 2 s J 

milligram per square meter second. 

mg m~ ! s- 1 

milligram per square meter second. 

micromole (H : 0) per square centi- 
meter second, nmol cm" 2 s- 1 

milligram per square centimeter 
second, mg cm 2 s- 1 

milligram per square decimeter hour. 
mg dm- 2 h- 1 

degrees (angle), ° 

Electrical Conduui it I • « fi i md Magnetism 

a 1 ' millimho per centimeter, mmho ci 

Water Measurement 

cubic feet per second, ft 3 s ' 

U.S. gallons per minute, gal it 





centiniole per kilogram, cmol kg 

gram per kilogram, g kg 

milligram per kilogram, mg kg" 1 

milliequivalents per 100 grams, 

100 g- 1 
percent. ' i 
parts per million, ppm 

becquerel, Bq 

becquerel per kilogram, Bq kg- 

gray. Cy (absorbed close) 
sievert, Sv (equivalent dose) 

picocurie per gram, pCi g- 

rem (roentgen equivalent man) 

Plant Nutrient Conversion 

Elemental Oxide 




Soil Fertility Replenishment in 
Africa: An Investment in Natural 
Resource Capital 

Pedro A. Sanchez, Keith D. Shepherd, Meredith J. Soule, 
Frank M. Place, Roland J. Buresh, and Anne-Marie N. Izac 

International Centre for Research in Agra forestry 
Nairobi, Kenya 

A. Uzo Mokwunye 

United Nations University 
Institute for Natural Resources in Africa 

Fred R. Kwesiga 

International Centre for Research in A, 

I i / 

Chipata, Zambia 

Cyrus G. Ndiritu 

Kenya Agricultural Research Institute 
Nairobi, Kenya 

University of Nairobi 
Nairobi, Kenya 


Soii-feriility depletion in smallholder farms is the fundamental biophysical root 
cause for declining per capita food production in sub-Saharan Africa. An average of 660 
kg N ha" 1 , 75 kg P ha" 1 , and 450 kg K ha" 1 has been lost during the last 30 yr from about 
200 million ha of cultivated land in 37 African countries. We propose an alternative 
approach, the replenishment of soil fertility as an investment in natural resource capital. 


This approach combines basic principles of soil science with environmental e 
Combinations of P fertilizers and organic inputs can replenish soil N and P nutrient stocks 
in Africa and restore service flows to near original levels. Phosphorus replenishment 
strategies are mainly mineral-fertilizer based, with biological supplementation. Nitrogen 
replenishment strati p are m rinl) biologically based with mineral-fertilizer supplemen- 
tation. Africa has ample phosphate rock (PR) deposits that can be either used directly or 
processed to reverse P depletion. Decomposing organic inputs may facilitate the use of PR 
in P-depleted soils. Leguminous tree fallows and herbaceous cover crops grown in situ 
play a major role in N capture and internal cycling in ways compatible with farmer con- 
straints. Soil-fertility replenishment was found profitable in three case studies, but small- 
holder farmers lack the capital and access to credit to make the initial investment. A cost- 
shared initial capital investment to purchase P fertilizer and germplasm for growth of 
organic inputs combined with effective microcredit for recurring costs such as fertilizers 
and hybrid seed is seen as the way forward. 

The continued threat to the world's land resources is exacerbated by the need to 
reduce poverty and unsustainable farming practices. During the last decade, food 
security was not a global priority, but studies such as the 2020 Vision (IFPRI, 
1995, 1996) and the World Food Summit (FAO, 1996) have shown that food 
security is one of the main global concerns as we move into the next century. 
Food insecurity encompasses food scarcity as well as the inability to purchase 
food, a poverty-related issue. Although food insecurity occurs throughout the 
developing world, it is most acute in sub-Saharan Africa (hereafter referred to as 
Africa), where the attainment of food security is intrinsically linked with revers- 
ing agricultural stagnation, safeguarding the natural resource base, and reducing 
population growth rates (Cleaver & Schreiber, 1994). 

In contrast to sustained increases in other parts of the developing world, per 
capita food production continues to decrease in Africa (Fig. 1-1; World Bank, 
1996a). This is largely a result of continuing rapid population growth, the high- 

n Asia and Lalin America (from World 


est of any region in the world, and rapid land depletion (Badiane & Delgado, 
1995; World Bank, 1995). In addition, about one-half of Africa's population is 
classified as absolute poor (those subsisting on per capita incomes of <1 U.S. dol- 
lar per day), and Africa has the highest proportion of undernourished children. To 
reverse this situation by the year 2020, Africa needs an annual, sustained growth 
rate in agricultural production of 4% (Badiane & Delgado, 1995). 

Three requirements for increasing per capita agricultural production have 
been identified: (i) an enabling policy environment for the smallholder farming 
sector (improved road infrastructure, access to education, credit, inputs, markets, 
and extension services), (ii) reversing soil-fertility depletion, and (iii) intensify- 
ing and diversifying land use with high-value products (Sanchez & Leakey, 

The purpose of this chapter is to propose a conceptual approach to increas- 
ing food security and poverty alleviation in Africa, the replenishment of soil fer- 
tility as tit! investment in natural resource capital, and to outline the elements of 
strategies for its implementation. This approach has been developed in response 
to a call by former World Bank President Lewis Preston for ideas on how to 
invest in natural resources in Africa (remarks to CGIAR Center Directors during 
International Centers' Week, 24 October 1994, Washington, DC). 

For some time, the research community has recognized low soil fertility, 
particularly N and P deficiencies, as one of the major biophysical constraints 
affecting African agriculture (Nye & Greenland, 1960; FAO, 1971; DeWit, 1981, 
1992; Penning de Vries & Djiteye, 1982; Mokwunye & Vlek, 1986; Pieri, 1989; 
Mokwunye, 1991; Yates & Kiss, 1992; Van der Pol, 1992; Aune, 1993; Smaling, 
1993; Wang'ati & Kebaara, 1993; Mokwunye etal, 1996). However, soil fertili- 
ty in Africa has seldom been considered a critical issue by the development com- 
munity, who until very recently have focused primarily on other biophysical con- 
straints such as soil erosion, droughts, and the need for improved crop germplasm 
(Lele, 1981; Eicher, 1982; Davis & Schirmer, 1987; World Bank, 1989, 1995; 
Anderson, 1994; Crosson & Anderson, 1995). Based on Smaling's (1993) nutri- 
ent balance studies and our own field observations across Africa, we have con- 
cluded that soil-fertility depletion in smallholder farms is the fundamental bio- 
physical root cause of declining per capita fond production in Africa, and soil- 
fertility replenishment should he considered as an investment in natural resource 
capital (Sanchez et al., 1996). 

By fundamental root cause, we mean that no matter how effectively other 
conditions are remedied, per capita food production in Africa will continue to 
decrease unless soil-fertility depletion is effectively addressed. During in the 
1960s, the fundamental root cause of declining per capita food production was 
the lack of short-statured, high-yielding varieties of rice {Oryza sativa L.) and 
wheat (Triticum aestivum L.) in Asia. Food security was only effectively 
addressed with the advent of improved germplasm in this region. Then other key 
aspects mat had been largely ineffective (irrigation, seed production, fertilizer 
use, pest management, research and extension services, and enabling government 
policies) came into play in support of the new varieties in Asia. The need for soil- 
fertility replenishment in Africa, therefore, is analogous to the need for Green 
Revolution-type germplasm in Asia three decades ago. 


Magnitude of the Problem 

The magnitude of nutrient depletion in Africa's agricultural land is enor- 
mous. Calculations from Smaling's seminal work (Stoorvogel & Smaling, 1990; 
Smaling 1993; Smaling et al, 1997, this publication) indicate that an average of 
660 kg N ha" 1 , 75 kg P ha" 1 , and 450 kg K ha- 1 has been lost during the last 30 
yr from about 200 million ha of cultivated land in 37 African countries, exclud- 
ing South Africa. This is equivalent to 1.4 t urea ha" 1 , 375 kg triple superphos- 
phate (TSP) ha" 1 or 0.9 t PR of average composition ha" 1 , and 896 kg KC1 ha- 1 
during the last three decades. These figures represent the balance between nutri- 
ent inputs as fertilizer, manure, atmospheric deposition, biological N2 fixation 
(BNF), and sedimentation, and nutrient outputs as harvested products, crop- 
residue removals, leaching, gaseous losses, surface runoff, and erosion. These 
values are the aggregate of a wide variety of land-use systems, crops, and agroe- 
cological zones in each country (Stoorvogel & Smaling, 1990). 

Africa is now losing 4.4 million t N, 0.5 million t P, and 3 million t K every 
year from its cultivated land. These rates are several times higher than Africa's 
annual fertilizer consumption, excluding South Africa — 0.8 million tN, 0.26 mil- 
lion tP, and 0.2 million tK (FAO, 1995). 

Commercial farms in the temperate region have averaged net positive nutri- 
ent balances in the order of 2000 kg N ha- 1 , 700 kg P ha- 1 , and 1000 kg K ha- 1 
during the last 30 yr in about 300 million ha of cultivated land, sometimes result- 
ing in groundwater and stream pollution (Frissel, 1978, Sanchez, 1994). Nutrient 
depletion in Africa, therefore, contrasts sharply with nutrient accumulation in 
temperate regions. 


How has such gross depletion of soil nutrients come about in Africa? 
Everywhere in the world people settle first in high-potential areas with fertile 
soils, adequate rainfall, and mild temperatures (Sanchez & Buol, 1975), such as 
parts of the highlands of eastern and Central Africa, the plateau of southern 
Africa, and some river basins in West Africa. The Lake Victoria Basin in East 
Africa is one example, and now supports one of the densest rural populations in 
the world, 500 to 1200 inhabitants km" 2 (Hoekstra & Corbett, 1995). 

Such settlements were first supported by the originally high soil fertility. As 
populations grew, this fertility was gradually depleted by crop-harvest removals, 
leaching, and soil erosion, when farmers were unable to sufficiently compensate 
these losses by returning nutrients to tire soil via crop residues, manures, and min- 
eral fertilizers (Shepherd & Soule, 1998). 

Smallholder farmers also cultivate low-potential areas primarily in subhu- 
mid and semiarid areas, where many of the sandy soils are naturally infertile. 
Still, the smaller soil nutrient stock is also being depleted in these areas (Pieri, 
1989; Smaling, 1993; Sanders etal., 1996). 

Two overarching reasons for the nutrient depletion process are (i) the 
breakdown of traditional practices and (ii) the low priority given to the rural sec- 


tor. Increasing pressures on agricultural land have resulted in much higher nutri- 
ent outflows and the subsequent breakdown of many traditional soil-fertility 
maintenance strategies, such as fallowing land, intercropping cereals with legume 
crops, mixed crop-livestock farming, and opening new lands. Such strategies 
have not been replaced by an effective fertilizer supply and distribution system 
(Sanders et al., 1996). Traditional African coping strategies were not capable of 
adjusting quickly enough to rapid population growth combined with decreasing 
farm size, soil fertility, and fuelwood availability (Cleaver & Schreiber, 1994). 
Nevertheless, African farmers repeatedly outperform the weather; current crop 
yields over time fluctuate considerably less than indices of rainfall (Dommen, 
1988). Continued population pressure has reduced farm sizes to the point where 
farms can only provide adequate living for their families if the land is farmed 
very intensively and if there is off-farm income. 

Most African governments have used agriculture as a main source of rev- 
enue by restricting producer prices and taxing exports (Cleaver, 1993). Relatively 
little attention has been paid to rural areas by national governments, lowering the 
relative returns to farming. Therefore, few improved soil-fertility management 
technologies have been widely accepted by African smallholder farmers 
(Conway & Barbier, 1990). The low national priority given to the rural sector in 
Africa also results in poor road and market infrastructure, lack of timely access 
to credit and inputs at reasonable cost, lack of timely information, and ineffective 
extension systems (Badiane & Delgado, 1995; Tomich et al., 1995). Soil-fertility 
depletion, therefore is largely a consequence of socioeconomic constraints and 
policy distortions. 

Because the soil resource has not kept its productive capability over time, 
farmers have witnessed low and declining yields. Current crop yields are low due 
to poor agronomic practices, droughts, weed and pest attacks, lack of cash for 
investment, and soil-fertility depletion. Several decades of nutrient depletion 
have transformed originally fertile lands that yielded 2 to 4 t ha" 1 of cereal grain, 
into infertile ones where cereal crops yields of <1 t ha" 1 are common. For exam- 
ple, a long-term trial in Kabete, Kenya, indicates that a fertile, red soil (Oxic 
Rhodudalf) lost about 1 t ha' 1 of soil organic N and 100 kg P ha- 1 of soil organ- 
ic P during 18 yr of continuous maize (Zea mays L.)-common bean (Phaseolus 
vulgaris L.) rotation in the absence of nutrient inputs. Maize yields without N and 
P fertilizer inputs decreased from 3 to 11 ha" 1 during that period (Qureshi, 1991; 
Swift etal., 1994; Kapkiyai, 1996; Bekunda et al., 1997, this publication). 

Nutrient depletion rates are field specific, depending on the way each par- 
ticular field has been managed over decades. This results in a mosaic of degrees 
of nutrient mining at the landscape scale. At the national scale, there are areas that 
have not suffered much from nutrient depletion, because of the low intensity of 
use or the use of fertilizer for export crops. Still, nutrient-depleted, smallholder 
farms in Africa are much more common that ones where this constraint is not a 
major problem. 

Nutrient depletion rates vary with soil properties. The proportion of nutri- 
ents lost is normally greater in sandy soils, but the total nutrient loss is greater in 
clayey soils. This is largely because soil organic matter (SOM) particles are less 
protected from microbial decomposition in sandier soils than in loamy or clayey 


ones (Sanchez, 1976; Pieri, 1989; Swift etal., 1994). This is one major difference 
between the nutrient-depleted, high-potential areas of East and southern Africa 
with predominantly loamy and clayey soils, and semiarid areas of West, East, and 
southern Africa with predominantly sandy soils. 

The end result is mat soils have deteriorated significantly, especially in 
terms of P levels and SOM. It now requires a major investment to restore soils to 
a sufficient level of fertility for sustainable crop production. 

Consequences of Nutrient Depletion 

On-Farm Effects 

A marked decline in crop productivity and food security are the main con- 
sequences of the policies that result in soil-fertility depletion in Africa. Nutrient 
depletion per se also produces negative on-farm side effects and exacerbates sev- 
eral off-farm effects or externalities. On-farm effects include less fodder for cat- 
tle, less fuelwood for cooking, and less crop residues and cattle manure to recy- 
cle nutrients. These effects often increase runoff and erosion losses because there 
is less plant cover to protect the soils. In sandy soils, the topsoil structure may 
collapse resulting in soil compaction or surface sealing. 

Economic and Social Externalities 

The negative effects of soil nutrient depletion extend beyond farming 
households into the community, regional, and national scales. Soil nutrient deple- 
tion lowers the returns to agricultural investment, which reduces nonfarm 
incomes at the community level through multiplier effects (Delgado et al, 1994). 
Other consequences of depletion are decreased food security through lower pro- 
duction and resulting higher food prices, increased government expenditures on 
health, more famine relief, and reduced government revenue due to less taxes col- 
lected on agricultural goods. 

Perhaps the most important negative social externality of soil-fertility 
depletion is its link to lower employment and increased poverty. The vast major- 
ity of the poor live in rural areas in the tropics (World Bank, 1990). As long as 
returns to agriculture are limited by nutrient depletion, farm employment and 
spillover nonfarm employment opportunities will remain low, sustaining severe 
poverty. But these externalities are not confined to rural communities, as poverty 
often pushes individuals and households into urban areas. The influx of rural 
migrants puts a greater strain on the limited urban infrastructure; and unemploy- 
ment, crime, and political unrest sometimes result (Homer-Dixon et al., 1993). 
This situation is typical in high-potential areas of eastern and southern Africa — 
particularly in Burundi, Ethiopia, Kenya, Malawi, Rwanda, Tanzania, Uganda, 
and Zambia as well as in low- and high-potential areas of West Africa. 

Environmental Externalities 

Soil-fertility depletion also exacerbates several environmental problems at 
the national and global scales. Increased soil erosion, particularly in steep areas, 


causes more unwanted sedimentation, siltation of reservoirs and of coastal areas, 
and in some cases eutrophication of rivers and lakes. There is evidence of these 
processes occurring in some African rivers and lakes (Melack & Maclntyre, 
1992), including Lake Victoria, where erosion from surrounding nutrient-deplet- 
ed lands is widespread. 

The loss of topsoil organic C associated with soil nutrient depletion results 
in additional CO2 emissions to the atmosphere from decreasing soil and plant C 
stocks. Assuming a C/N ratio of 10:1 in SOM, the average N depletion rate of 22 
kg N ha-' yr ' represents an average rate of C loss of 220 kg C ha" ' yr ' from 200 
million ha of cultivated soils of Africa. Assuming that 40% of the losses are due 
to erosion (Smaling, 1993) and that C in sediments resulting from erosion is not 
lost to the atmosphere as CO 2, 27 Tg (10 12 g) of C are annually emitted to the 
atmosphere from cultivated land in Africa. This amounts to about 2 1 % of the 130 
Tg C annually emitted from land degradation (Lai et al, 1995). 

Carbon loss is a reversible process in soils as long as their clay contents are 
not decreased by erosion. Equilibrium soil organic C content is determined by 
soil temperature, moisture, clay content, and the balance between C input and 
decomposition rates (Nye & Greenland, 1960; Sanchez, 1976). It has been possi- 
ble to increase organic C to their original levels by sound fertilizer and residue 
management practices in the temperate regions (Buol et al, 1990; Buol & Stokes, 
1997, this publication). This C sequestration process is gradual (Giller et al., 
1997, this publication) and definitely not instantaneous. Although it takes place 
primarily in the topsoil, it also can occur in the subsoil when deep-rooted grass- 
es and trees are introduced in degraded lands (Fisher et al, 1994; Sanchez, 1995). 
The cry for increasing SOM in sandy soils, so often heard in West Africa, can 
only occur in nutrient-depleted soils, but never up to the levels found in high- 
potential clayey soils. With these caveats in mind, decreased CO? emissions and 
increased C sequestration can be a positive environmental externality of replen- 
ishing soil fertility. 

Soil-fertility depletion decreases above- and belowground biodiversity and 
increases the encroachment of forests and woodlands in response to the need to 
clear additional land (Sanchez, 1995). This is particularly relevant to the Miombo 
woodlands of southern Africa and to the rainforest remnants in the Great Lakes 
region and in eastern Madagascar, both of which harbor unique animal biodiver- 


Existing Approaches 

Traditional approaches for soil-fertility management range from recurring 
fertilizer applications to low external input agriculture based on organic sources 
of nutrients. Although both extremes work well in specific circumstances, they 
pose major limitations for most smallholder farmers in Africa. 

Recurring Fertilizer Applications 

Fertilizer use is the obvious way to overcome soil-fertility deplet 

indeed it has been responsible for a large part of the sustained increases 


capita food production that have occurred in Asia, Latin America, and the tem- 
perate region, as well as in the commercial farm sector in Africa (Mokwunye & 
Vlek, 1986; Mokwunye & Hammond, 1992; Borlaug & Dowswell, 1994; Buol & 
Stokes, 1997, this publication). There is nothing wrong, biophysically or envi- 
ronmentally, with fertilizers when properly used. Fertilizers provide the same 
nutrients as organic sources to plants. Plants cannot distinguish nitrate or phos- 
phate ions they absorb from organic inputs from those they absorb from mineral 
fertilizer:, (hereafter referred to as inorganic fertilizers!. 

Most smallholder farmers in Africa appreciate the value of fertilizers, but 
they are seldom able to apply them at the recommended rates and at the appro- 
priate time because of high cost, lack of credit, delivery delays, and low and vari- 
able returns (Badiane & Delgado, 1995; Runge-Metzger, 1995; Heisey & 
Mwangi, 1996; Larson & Frisvold, 1996). Such constraints are largely due to the 
lack of an enabling policy environment in rural areas caused by the deficient road 
and market infrastructure typical in most African countries. 

The price of fertilizers in rural areas of Africa is usually at least twice the 
international price (Bumb & Baanante, 1996). Transport costs are about seven 
times higher in Africa than in the USA (U.S. dollars in t km" 1 ). Transport and 
other costs (import duties, demurrage, taxes) more than double the international 
price by the time fertilizers reach the farmer in Malawi (Donovan, 1996). This 
occurs in spite that during the past 25 yr the real international price of fertilizers 
has decreased by about 38% forN and >50% for P (Donovan, 1996). 

In the past, many African countries subsidized fertilizers, however, the 
removal of fertilizer subsidies by most African governments as part of the 
Structural Adjustment Programs in the last decade has tripled or quadrupled fer- 
tilizer prices in relation to crop prices in many African countries (Holden & 
Shanmugarathan, 1994; Bumb & Baanante, 1996). Furthermore, since fertilizer 
recommendations are normally formulated to cover broad areas with diverse 
soils, farmers also lack information about the best fertilizer to use for their par- 
ticular fields and cropping practices, making the crop response to fertilizers more 
erratic and less profitable. 

These policy and information constraints can certainly be overcome, there- 
by in the longer term resulting in increased food security and reducing poverty. 
An excellent example of a promising approach is the Sasakawa Global 2000 pro- 
ject in Ethiopia, where many policy distortions have been overcome (Quinones et 
al, 1997, this publication). 

Organic Farming 

The exclusive use of organic inputs as external nutrient sources has been 
advocated as a logical alternative to expensive fertilizers in Africa (Reijntjes et 
al, 1992). The main advantages of this approach are (i) the replacement of scarce 
or nonexistent capital for labor and (ii) the fact that cattle manures or green 
manures contain all essential nutrients plus C, the source of energy for soil biota 
that regulates nutrient cycling. 

One of the main arguments against the use of organic inputs is their low 
nutrient concentration in comparison with inorganic fertilizers. Animal manures 


and plant material contain from 1 to 4% N (10-40 g N kg" ) on a dry weight 
basis, while inorganic fertilizers contain from 20 to 46% N (200 — 460 g N kg" 1 ) 
and are already dry. To haul the 100 kg N generally needed for a 4 t ha" 1 maize 
crop, it would take 217 kg of urea or 20 1 of leaf biomass with 80% moisture and 
a 2.5% (25 g N kg" ' ) N concentration on a dry weight basis. Furthermore, organ- 
ic inputs are very low suppliers ofP because of their low concentrations (Palm, 
1995; Palm et al., 1997b, this publication). 

It takes soil fertility to grow organic inputs, be they green manures, litter- 
fall, plant biomass for transfer, composts, or animal manures. In nutrient-deplet- 
ed soils it is difficult to grow enough forage to feed cattle and produce sufficient 
quantities of manure. The nutrient content of manures varies widely with soils 
and fodder availability in Africa (Murwira et al, 1995; Probert et al., 1995). 

It is important to distinguish between nutrient inputs and nutrient cycling 
(Sanchez & Palm, 1996). At the field scale, nutrient inputs are additions from 
outside the system, such as No fixed from the air by legumes or inorganic fertil- 
izers. Cattle manures are inputs if the manure was produced from forage grown 
outside the field. 

Nutrient cycling refers to the transfer of nutrients already in the field from 
one component to another, for example (i) the return of maize stover back to the 
soil, (ii) cattle manure and urine deposited by animals while grazing crop 
residues, and (iii) the transfer of nutrients from trees to the soil through prunnings, 
leaf drop, or root decomposition in agroforestry systems. Nutrient cycling is 
extremely important, but nutrient-depleted soils need inputs from outside the 

African farmers, in our view, do not need low-input, low-output systems 
that do not address poverty alleviation. Given the limitations of the extremes of 
either pure organic inputs or pure inorganic inputs, the authors feel it is time for 
a more robust approach that provides fresh alternatives and increases the basket 
of options available to farmers and policy makers. 

Natural Resource Capital 

Capital, or the basis of value, may be divided into four general categories: 
natural, manufactured, human, and social capital (Serageldin, 1995). Natural cap- 
ital is the stock of environmentally provided assets, such as the atmosphere, soil, 
water, forests, fish, wildlife, and wetlands. Manufactured capital is the capital 
people usually consider in financial and economic terms (e.g., money, houses, 
roads, factories, and vehicles). Human capital represents the education, health, 
and nutrition of individuals, while social capital reflects the institutional and cul- 
tural basis from which a society functions. The flow of goods and services from 
natural capital can be renewable or nonrenewable, and marketed or nonmarketed 
(Serageldin, 1995). 

In the past, natural capital and its decline has seldom been taken into 
account. Now that human expansion has made large claims on the environment, 
natural capital has become an important limiting factor to economic develop- 
ment. For example, the loss of forests or fish populations has a greater impact on 


the possibilities for long-term, sustainable economic development than the loss of 
manufactured capital such as sawmills or fishing boats (Serageldin, 1995). 

The current situation of depleted natural capital in many countries reflects 
decades of natural capital extraction in an attempt to build up other types of cap- 
ital. Such a strategy was intended to lead to demand-driven investments across all 
types of capital and ultimately to sustainable growth; however, investments in 
natural resources remain low in most African countries. In many of them, natur- 
al resource capital values have depreciated to extremely low levels. 

In the longer term, sustainable growth and poverty alleviation will depend 
on investment in and improvement of all four types of capital. There is an urgent 
need to reexamine past strategies and consider investments in natural capital as a 
means to achieving this goal. In the impoverished rural areas of Africa, encour- 
aging investment in land resources may not only have the greatest impact on 
poverty alleviation in the short-term, but it may act as a catalyst for broader-based 
capital investment and growth in the long term. The remainder of this chapter 
focuses on the potential for investing in one type of natural resource capital — soil 

Nutrient Capital 

Soil fertility is an important form of renewable natural capital. Plant nutri- 
ents in the soil, however, are among the least resilient components of sustain- 
ability (Fresco & Kroonenberg, 1992). The maintenance of soil fertility involves 
the return to the soil of the nutrients removed from it by harvests, runoff, erosion, 
leaching, and other loss pathways (Aune, 1993). 

The soil nutrient capital of Africa is being mined, just like mineral deposits 
of metals or fossil fuels. As long as poverty and population pressure hinder farm- 
ers from replenishing this lost capital, the service flows emanating from nutrient 
capital will inevitably decrease. These service flows include on-farm soil fertili- 
ty and crop production (valued by individual farmers), food security and soil con- 
servation (valued by national societies and farmers), poverty alleviation, 
enhanced C sequestration, and biodiversity conservation (valued by the global 
society and future generations). 

At the farm scale, it makes sense for farmers to reverse the mining of their 
nutrient capital up to the point where the marginal costs of nutrient replenishment 
are covered by the marginal benefits. For annual fertilizer applications, this 
means that the value of increased production brought about by the fertilizer 
should be sufficiently high to cover costs of the fertilizer, compensate for risk, 
and provide a reasonable return to the farmer. 

Fertilizer application by smallholder farmers to food crops is often not prof- 
itable due to the combination of high fertilizer prices discussed earlier, low prices 
for food crops, and high risk. Even when fertilizer applications are profitable, 
many farmers cannot afford to purchase fertilizer at the beginning of the season 
when other, more basic needs, are pressing. Functioning credit markets could 
alleviate this constraint, but they have been difficult to develop for small farmers 
growing food crops in Africa due to the risky nature of rainfed agriculture, the 
lack of collateral, and the high cost of administering small loans to many small- 


holder farmers. On the other hand, fertilizers are often applied to cash crops such 
as coffee (Cqffea sp.) and tea [Camellia sinensis (L.) Kuntze] where the returns 
to fertilization are high and credit is available through cooperatives. 

Nutrient Capital Defined 

Nutrient capital can be defined as the stocks of N, P, and other essential ele- 
ments in the soil that become available to plants during a time scale of 5 to 10 yr 
(Sanchez & Palm, 1996). The two most widespread limiting nutrients to food pro- 
duction in Africa are N and P, in that order (Ssali et al, 1986; Woomer & 
Muchena, 1996; Bekunda et al., 1997, this publication). For example, in a series 
of fertilizer trials conducted throughout the Kenyan highlands, N and P deficien- 
cies were reported in 57 and 26% of the cases, respectively (Kenya Agricultural 
Research Institute, 1994). Potassium, Ca, Mg, S, and micronutrient deficiencies 
and Al toxicity do occur in specific circumstances in Africa, but not to the extent 
ofN and P deficiencies. Potassium depletion rates (15 kg K ha" 1 yr" 1 ) are six 
times that of P (Smaling et al., 1997, this publication), but crop responses to K 
fertilization are rare in Africa, except in sandy savanna soils (Ssali et al, 1986). 
This is probably due to the high K capital in many other parts of Africa and the 
low demands for K due to the current low crop yield levels. Consequently, in this 
chapter we focus on the two main limiting nutrients. 

We propose to define N capital as the labile pools of soil organic N that 
seem to be well correlated with N release rates, such as particulate organic N 
(Kapkiyai, 1996), and N in the light fraction of SOM (Barrios et al., 1996). We 
also propose that P capital be defined as the labile pools of soil organic P (such 
as particulate organic P) together with sorbed or fixed P on Fe and Al oxides and 
hydroxides at the surface of layer-silicate clay particles. Nutrient capital may be 
expressed as kilograms per hectare of N or P within the rooting depth of plants. 
Precise methodologies, however, need to be developed. 

Capital Stocks and Service Flows 

There is an exact congruence between the concepts of capital stocks and 
service flows in economics and that of nutrient pools and fluxes in soil science 
(J.K. Lynam, 1997, personal communication). The above-defined nutrient capital 
stocks as discrete pools fit well with economic concepts. Nutrient fluxes during 
the growing period are synonymous to service flows. Such fluxes subtract from 
the nutrient capital and are thus analogous to the concept of depreciation. 

Our goal is not to maximize nutrient stocks in the soil but rather to deter- 
mine the minimal size of the nutrient stock that will maximize service flows, or 
the value of crop production for several years. An important question for P 
replenishment is whether service flows will be maximized with a one-time appli- 
cation of P or by gradually increasing the P stock through annual applications. In 
the case of P-depleted, high P-sorbing soils, small annual fertilizer P applications 
will go primarily to slowly available pools, making little P available to plants, 
therefore minimizing service flows. This calls for a strategy to rapidly replenish 
capital P pools, rather than less efficient, gradual build ups. Key advantages of 
rapid P replenishment are economies of scale and cost savings in delivering and 



applying P, particularly PR. The downsides are lack of credit and capital by farm- 
ers and the discount rate. As the discount rate increases, higher service flows 
from the investment are required to pay for the investment. In the case of N, how- 
ever, the size of the capital N stocks cannot be built instantaneously like P capi- 
tal stocks, so gradual build ups are needed. The critical factor is not the size of 
the N capital stocks, but the cycling rate (Giller et al., 1997, this publication). 

Nitrogen and P, therefore, behave differently in terms of their replenish- 
ment strategies. We will tackle P replenishment first, because although less geo- 
graphically widespread, it is the most critical of the two. 


Phosphorus deficiency is widely considered the main biophysical con- 
straint to food production in large areas of farmland in subhumid and semiarid 
Africa (Penning de Vries & Djiteye, 1982; Ssali et al., 1986; Bationo etal., 1986, 
1996). Phosphorus dynamics in soils are complex, because they involve both 
chemical and biological processes and the long-term effects of sorption (fixation) 
and desorption (release) processes. The low concentration and low solubility of P 
in soils frequently make P a limiting factor. The main features of the P cycle are 
shown in Fig. 1-2. The following section highlights the concepts and processes 


Phosphorus inputs to farmer fields in Africa consist primarily of inorganic 
fertilizers and organic sources such as biomass, manures, and composts gathered 

farms of Africa. 


from outside the field. The P content of plant residues and manures is normally 
insufficient to meet crop requirements. Plant materials applied as organic inputs 
contain 8 to 12 kg P ha" 1 when applied at the top realistic rate of 4 t dry matter 
ha" 1 (Palm, 1995). This is about one-half the P requirements of a 4 t ha' 1 maize 
grain crop, which accumulates about 18 kg P ha- 1 in its tissues (Sanchez, 1976). 
Phosphorus, unlike N, is not captured from the atmosphere by biological fixation 
nor from deep in the soil profile, due to the very low concentrations of available 
P in the subsoil and low root-length densities (International Atomic Energy 
Agency, 1975). Consequently, P fertilizers are almost always necessary to over- 
come P depletion (Breman, 1990; Mclntire & Powell, 1995). 

The gathering of green plant material from boundaries or adjacent fields 
and their addition to another field is known as biomass transfer. Most of the bio- 
mass transfers practiced by African farmers consists of leguminous plants and 
grasses. There is increasing evidence, however, that some nonleguminous shrubs 
may accumulate higher than normal concentrations of P in their biomass than 
legumes. Tithonia [Tithonia diversifolia (Hemsley) A. Gray], a common farm 
hedge species native of Mexico and found at middle elevations throughout trop- 
ical Africa (Niang et al., 1996; Palm et al., 1997a) and southeast Asia (Nagarajah 
& Amarasiri, 1977; Nagarajah & Nizar, 1982; Cairns & Garrity, 1998) has unusu- 
ally high concentrations of P (0.27-0.38% P or 2.7-3.8 g kg" 1 ) in its leaf biomass 
(Gachengo, 1996; Nziguheba et al, 1998). These levels are far superior to those 
in commonly used legumes in agroforestry and as herbaceous cover crops, which 
range in the order of 0.15 to 0.20% P (1.5-2.0 g kg" 1 ; Palm, 1995). Reasons for 
such high concentrations remain speculative but members of the Asteraceae 
(Compositae) family, to which tithonia belongs, are effective nutrient scavengers 
(Szottetal., 1991; Garrity & Mercado, 1994). 

Internal Flows 

Plants absorb phosphate ions, which come into the soil solution through the 
dissolution of P-bearing weatherable minerals, the dissolution of P fertilizers, the 
desorption of sorbed P, and the mineralization of soil organic P. Mycorrhizal asso- 
ciations facilitate the capture of phosphate ions by effectively enhancing the vol- 
ume of soil exploited by roots and the efficiency with which P is extracted (Lajtha 
& Harrison, 1995). Mycorrhizae are key facilitators for P capture by plants in P- 
depleted soils (Tinker, 1975; Hedley et al., 1995). 


The dissolution of P from the main P-bearing weatherable mineral in soils 
is only important in areas of Africa with soil derived from young alluvium or 
basic rocks high in P content. This is the case with many Vertisols and some cal- 
careous and alluvial soils; however, weathering and rapid nutrient depletion have 
exhausted this source in most of the intensively cultivated lands of Africa. In soils 
derived from granitic or basaltic materials such as the bulk of Alfisols of the 
African subhumid and semiarid areas, the P reserves in weatherable minerals 
have always been low. 



Phosphate ions may be taken away from the soil solution and sorbed by Fe 
and Al oxides and hydroxides at the surface of clay particles. While these sorbed 
phosphate ions are unavailable to plants in the short run, they are slowly desorbed 
and released to the soil solution during a period of several years. Phosphorus 
sorption is considered the most important process controlling P availability in 
soils (Lajtha & Harrison, 1995) and has long been considered a major constraint 
to crop production (Sanchez & Uehara, 1980). Current systems thinking on sus- 
tainability, however, turns P sorption from a liability into an asset. Large appli- 
cations of P fertilizers could become P capital as sorbed P. Subsequent P desorp- 
tion is the service flow. 

Phosphate sorption is controlled by clay surfaces and is only important in 
the topsoil where P fertilizers are applied. High P-sorbing soils, therefore, can be 
identified as those with clayey topsoils having red colors indicative of high con- 
tents of Fe and Al hydrous oxides, usually accompanied by a strong granular 
structure (Sanchez et al., 1982). These can be collectively termed oxidic soils, 

Fig. 1-3. ICRAF estimates a total of 530 million ha of P-fixing soils in Africa, shown in dark si 


and are mainly classified as Oxisols, clayey Ultisols, rhodic, or oxic groups or 
subgroups of clayey Alfisols, and Inceptisols in soil taxonomy (Soil Survey Staff, 
1992). In the FAO Legend they are classified as Nitisols, clayey Ferralsols, and 
clayey Acrisols (FAO, 1988). ICRAF estimates there are about 530 million ha of 
high P-sorbing soils in Africa, which represents 25% of tropical Africa's land 
area (Fig. 1-3). 

It is important to recognize that we are excluding from consideration two 
other P-retention processes — those found in volcanic and in calcareous soils. P 
sorption in volcanic soils (Andisols) occurs in minerals collectively known as 
allophane. Upon P fixation these minerals actually open new P-sorbing sites, 
making them an ultimate sink for P with very slow desorption rates (Sanchez et 
al., 1982; Frossard et al., 1995). Therefore the capital stocks build up, but the ser- 
vice flows are minimal. Fortunately there are not many P-depleted volcanic soils 
in Africa, although these soils are locally important. Phosphorus also is retained 
in calcareous soils by precipitation with calcium carbonates, which results in the 
formation of calcium phosphates — apatites. The release of P from apatite is a 
straight solubility reaction, which is slow in high pH, high Ca soils. The magni- 
tude of P fixation in calcareous soils is small, hence the growth in capital stock is 
small. The service flows are also small due to the low solubility of the product 
(Frossard et al., 1995). Consequently P-replenishment strategies are limited to 
soils with hydrous oxides of Fe and Al, which represent the bulk of smallholder 
farms in Africa. 


Most of the P sorbed by Fe and Al compounds is slowly released back to 
the soil solution, providing service flows for 5 to 10 yr. For example, 68% of a 
150 kg P ha" 1 application of diammonium phosphate (DAP) was recovered by a 
maize-bean rotation over 5 yr in an Oxisol of western Kenya (Heinemann, 1996). 
The residual effect of large phosphate applications is due to the desorption 
process, and in the Cerrado region of Brazil the duration of the residual effect 
increases with increasing rates of P application (Goedert, 1985). 


Plants convert inorganic P absorbed from the soil solution into organic 
forms in their tissues. The addition of plant material grown in situ to the soil as 
litterfall, root decay, green manure incorporation, crop-residue returns, and ani- 
mal excreta (in grazing systems), and its subsequent decomposition results in the 
formation of organic forms of soil P. Microbes assimilate phosphate ions in the 
soil solution into organic forms in their biomass, a process referred to a P immo- 
bilization. Mineralization of soil organic P, including recently immobilized bio- 
mass P, releases it once again to soil solution P, which is readily available to 
plants, thus providing an additional service flow. This is the process of organic 

Many trees, shrubs, and important crop species have the ability to exude 
organic acids from their roots or have mycorrhizal associations that help dissolve 
inorganic soil phosphates not otherwise available to crops (Lajtha & Harrison, 


1995). Pigeonpea [Cajanus cajan (L.) Millsp.] secretes pisidic acid in calcareous 
soils (Ae et al, 1990) increasing the plant's P uptake. Inga (Inga edulis Mart.) is 
believed to have access to P not available to maize and beans (Hands et al, 1995). 
Pigeonpea and inga are legumes, which are known to acidify their rhizosphere in 
the process of N2 fixation. In such cases, organic cycling has the advantage of 
transforming essentially unavailable forms of inorganic soil P into more available 
organic forms. Phosphorus depletion, however, can have serious negative effects 
on N cycling because adenosine triphosphate is needed in larger quantities for N2 
fixation by legumes than by plants that do not fix N (Giller & Wilson, 1991). 


The two main P loss pathways in Africa are crop-harvest removals and soil 
erosion (Smaling, 1993). Most of the P in cereal crops and grain legumes is accu- 
mulated in the grain and removed from the field at harvest. The proportion of P 
cycled back to the soil in grain crops, assuming complete crop residue return, is 
in the order of 40%, in contrast with about 50 to 70% for N and 90% for K 
(Sanchez, 1976; Sanchez & Benites, 1987). On smallholder farms in Africa, most 
crop residues are not returned to the field where they were produced because they 
are used for cattle fodder, fencing, or cooking fuel. This results in 100% removal 
of the P accumulated by crops for human nutrition. 

While grain-harvest removal is a desirable outcome, soil erosion is envi- 
ronmentally dangerous since a P-enriched topsoil, if eroded, can cause eutrophi- 
cation of surface waters (Sharpley et al., 1995). Losses of P by leaching are rare, 
except in very sandy soils, such as those in the Sahel (Brouwer & Powell, 1995). 

Replenishing Phosphorus Capital 

Replenishing P capital can only be accomplished with P fertilizer inputs. 
Experiences in the high P-sorbing soils of the Cerrado region of Brazil during the 
last 30 yr have shown that large applications of P can replenish P and the resid- 
ual effect of such replenishment lasts for at 5 to 10 yr (Lopes, 1983, 1996; 
Goedert, 1985, 1987; Lopes & Guilherme, 1994). These corrective applications, 
along with subsequent maintenance applications, sound agronomic practices, and 
an enabling policy environment revolutionized farming in the Cerrado, which is 
now a major food-exporting region (Ableson & Rowe, 1987; Sanchez, 1994; 
Lopes 1996). Farmers who used P as a capital investment in the Brazilian 
Cerrado, along with annual applications of lime and other inputs, achieved inter- 
nal rates of return on their investment of 96% (Ribeiro, 1979). This experience 
lends support to this concept for Africa. 

Strategic options include a one-time application of P versus repeated small- 
er applications, and the use of PR vs. soluble P fertilizers. These options are 
directly related to two key soil properties: P sorption capacity and soil acidity. A 
one time corrective application in the order of 150 to 500 kg P ha" 1 of TSP is 
probably the most straightforward way to replenish P capital and the effect lasts 


for several years in high P-sorbing soils (Yost et al., 1979; Goedert, 1987). High- 
reactive PR can be used in acid soils with similar agronomic efficiencies as super- 
phosphates, but seldom in soils with pH values above 6.2 (Goedert & Lobato, 
1980; Yost etal., 1982; Goedert, 1985; Frossard et al, 1995; Buresh et al., 1997, 
this publication). In sandy soils of the Sahel, where both P sorption and P avail- 
ability are extremely low, medium -reactive PRs applied at low rates (15 to 30 kg 
P ha" 1 ) have relative agronomic efficiencies of 68 to 104% within a 3-yr period 
(Bationo & Mokwunye, 1991). 

Phosphate Fertilizer Sources 

Sub-Saharan Africa has vast quantities of PR deposits of varying quality, 
some of which are of sedimentary origin and reactive (McClellan & Notholt, 
1986; Van Kauwenbergh etal, 1991; Buresh etal., 1997, this publication). High- 
to medium-reactive (>15 g citrate-soluble P kg" 1 ), sedimentary PR deposits are 
found in Angola, Burkina Faso, Mali, Niger, Senegal, and Togo, while highly 
reactive biogenic PR deposits occur in Tanzania and Madagascar. Most of these 
materials have been found to be suitable for direct application, while others are 
only effective when partially acidulated, applied in combination with soluble P 
fertilizers, or mixed with organic inputs (Hammond et al., 1986; Bationo et al., 
1986, 1990; Bationo & Mokwunye, 1991; Buresh & Tian, 1997; Buresh et al., 
1997, this publication). Igneous PR deposits, found in Burundi, Congo, Kenya, 
South Africa, Uganda, Zambia, and Zimbabwe are seldom suitable for direct 
application and are best used to manufacture superphosphates or modified as par- 
tially acidulated or compacted PR (Bationo & Mokwunye, 1991; Buresh et al., 
1997, this publication). 

The choice of P source also depends on many other factors, such as cost dif- 
ferentials between PR and superphosphate, soil acidity, and the P sorption capac- 
ity of the soil. Cost differences per kg P can be major, with locally produced high- 
ly reactive PRs having a competitive advantage in some instances. The more acid 
the soil, the more rapid the dissolution rate of PR. The soil and the superphos- 
phate factory basically operate in the same way, dissolving PR with the addition 
of acids (Sanchez & Salinas, 1981). Furthermore, high P sorption enhances the 
dissolution of PR by reducing the concentration of P in solution around the PR 
particle (Smyth & Sanchez, 1982; Kirk & Nye, 1986). 

Residual Effects 

The duration of crop yield responses to P applications depends on the 
amount of P applied, the soil's P sorption, and cropping intensity. The larger the 
P application rate — the longer the residual effect. Low P-sorbing soils have short- 
er residual effects than high P-sorbing soils. The higher the number of crops har- 
vested per year — the shorter the residual effect. Replenishment strategies for high 
P-sorbing, clayey, red soils of East and southern Africa will therefore differ from 
strategies for low P-sorbing, sandy soils of the Sahel, where smaller and more 
frequent applications are required (Bationo et al, 1996). Given these variables, as 
well as logistical, financial, and infrastructure considerations, the choice of P fer- 
tilizer source and the rate used for replenishment is site and situation specific. 


Combining Inorganic and Organic Phosphorus Inputs 

One of the problems of P replenishment in Africa is that acidifying agents 
are likely to be needed to facilitate the dissolution of PR. Many P-depleted 
African soils have pH values above 6.2, which as stated earlier, are too high for 
rapid dissolution of reactive PR. The decomposition of organic inputs produces 
(i) organic acids that may help acidify PR or (ii) chelating agents that hind (a dis- 
solved from PR and thus stimulate the further dissolution of the PR. Mixing PR 
with compost has been shown to increase the availability of African PRs in at 
least two cases: Kodjari PR in Burkina Faso (Lompo, 1993) and Minjingu PR in 
Tanzania (Ikerra et al., 1994). Another option might is to mix finely-ground PR 
with poultry or cattle manure prior to application to nonacid soils. 

Organic anions produced during the decomposition of plant materials may 
temporarily reduce the P-fixation capacity of soils by binding to the Fe and Al 
oxides and hydroxides at surfaces of clay particles (Iyamuremye & Dick, 1996). 
Nziguheba et al. (1998) found that the rapid decomposition of 5.5 t ha~' of titho- 
nia dry biomass reduced the P sorption and increased the available P pools of an 
acid soil during a 16-wk period. This was attributed to the blocking of P sorption 
sites by organic anions produced during biomass decomposition. Through such a 
process, P availability and nutrient-use efficiency are temporarily increased dur- 
ing a grain crops' growth period. In contrast, no such effect has been observed 
with senna (Senna spectabilis DC; Gachengo, 1996). Indeed, little is known 
about the influence of organic materials on P solubilization and sorption-desorp- 
tion processes when organic materials are applied along with inorganic fertilizers 
(Palm et al., 1997b, this publication). 

On-farm research in western Kenya illustrates the potential on combining 
inorganic and organic sources of P in a moderate P-sorbing Oxisol with pH 5.1. 
Minjingu PR and TSP were the inorganic P sources applied either with 1.8 t ha" 1 
of tithonia dry biomass or with an equivalent N rate as urea (60 kg N ha" 1 ). 
Results of the first crop (Fig. 1-4) show that the application of tithonia without P 
fertilizer doubled maize yields in comparison with the equivalent N rate as urea. 

Fig. 1-4. Effect of combining tithonia biomass transfer (1.8 t c 
phate rock (PR) an i i s i ^ ' Li] I hate (1 >P) both applied a 
yield on an acid soil in western Kenya. Urea and tithonia were ; 
from ICRAF, 1997; Sanchez et al., 1997). 

ith Minjingu phos- 
250 kg P ha- 1 , on maize grain 
plied at 60 kg N ha- 1 (adapted 


The combination of tithonia biomass transfer with the replenishment rate of 250 
kg P ha" 1 as Minjingu PR increased maize yields fivefold (from 0.8 to 4.01 ha" 1 ). 
Tithonia combined with 250 kg P ha" 1 eliminated the differences in maize yield 
response between TSP and Minjingu PR (Sanchez et al., 1997; ICRAF, 1997). 
Maize crops in subsequent seasons are showing the same trends. The benefit from 
tithonia is partially attributed to addition of 60 kg K ha" 1 with the plant material. 
Subsequent research confirmed higher maize production with the sole application 
of tithonia biomass than with an equivalent rate of N-P-K as inorganic fertilizer 
in a soil deficient in N, P, and K (Bashir Jama, 1997, personal communication). 
Techniques to replenish soil P therefore, consist of P fertilization together 
with the effective use of organic sources. The integration of locally available 
organic resources with commercial P fertilizers may be the key to increasing and 
sustaining levels of P capital in smallholder African farms. 

Economic Considerations 

If P fertility is to be replenished for a number of years by a large, one-time 
capital investment, then the net present value (NPV) of the stream of net benefits 
generated by the P investment over a number of years should exceed zero at the 
chosen discount rate. In other words, the discounted value of the increased pro- 
duction, aggregated over the number of years the investment continues to provide 
benefits, must exceed the cost of the investment. 

Even if P replenishment meets this criterion and is privately profitable for 
the farmer, the credit market imperfections previously discussed, and the extreme 
poverty of many farmers will mitigate against the adoption of P replenishment 
without some form of outside assistance. The P replenishment strategy should be 
promoted mainly in regions that are known to have large areas of P deficient soils 
that have P-sorption capacity. In such areas, the aggregate private returns to P 
replenishment are likely to be the greatest benefit, but externalities generated by 
P replenishment at a wide scale provide additional societal benefits (poverty alle- 
viation, food security, employment effects, enhanced C sequestration, and biodi- 
versity conservation). Under conditions of missing capital and information mar- 
kets, assistance to farmers in replenishing P may be necessary in order for the 
potential private and social benefits of P replenishment to be realized. 


Phosphorus replenishment must usually be accompanied by N replenish- 
ment in order to be effective, because most P-deficient soils also are deficient in 
N; however, replenishing N stocks to near their original levels would require very 
large inputs of organic N. For example, an increase in soil organic N concentra- 
tion from 0.1 to 0.3% N (1-3 g N kg- 1 ) in the topsoil (0.2-m deep with a bulk 
density of 1.0 Mg m" 3 ) is equivalent to an application of about 160 t ha" 1 of dry 
biomass (2.5% N or 25 g N kg" 1 ) or 8700 kg ha"' of urea. Such large applications 
are clearly impractical and environmentally undesirable. In the short to medium 
term, increased soil N supply will depend on regular applications of N inputs. 


/NK^Crep offwto 

PtjretN Gain 




Uhwiliullon T" ^~) 

V ,n ° r ^ H { 

tonruUon 1 ( 

/y \ sohm*c |* 


Fiji. l-.i. Main features ot the N cycle using organic inputs in smallliokler farms of Africa. 

The gradual rebuilding of N capital is a worthwhile objective, however, to 
provide buffering against uncertainty in the farmers' ability to supply N inputs to 
every crop. As indicated by Giller et al. (1997, this publication) the main issue in 
N replenishment is not the size of the capital N stocks, but the cycling rate. 
Therefore, appropriate strategies are those that will provide sufficient levels of N 
inputs while crops are growing, and at the same time slowly rebuild N stocks. 

Nitrogen cycling consists of various inputs, outputs, and internal flows at 
the field scale (Fig. 1-5). Nitrogen capital stocks include the labile pools of soil 
organic N, such as microbial biomass N and N in the light fraction of SOM, 
which are correlated with N mineralization rates (Barrios et al., 1996; Kapkiyai, 
1996). Given the largely biological nature of the N cycle, organic inputs play a 
crucial role in N replenishment. 


Nitrogen inputs to a field consist mainly of inorganic fertilizers, biomass 
transfers, BNF, animal manures or composts produced outside the field, and 
nitrate capture from subsoil depths beyond the reach of crop roots. BNF becomes 
an input upon the conversion of atmospheric N2 gas into plant N by symbiotic 
plants followed by the addition of plant N to the soil. 

Inorganic fertilizers account for about one-third of the N inputs in Africa 
(Smaling, 1993), but they are used largely in mechanized agriculture and on 
export crops. Only three countries in sub-Saharan Africa — Nigeria, Zimbabwe, 
and South Africa — produce N fertilizers. Millions of smallholder farmers 
throughout Africa, however, use N fertilizers, most of which are imported. Heisey 
and Mwangi (1996) reported that 37% of the area planted to maize in 11 African 
countries received N fertilizers in the early 1990s. Because of the high price 


imported fertilizers at the farm gate and delays in delivery due to poor infra- 
structure (Donovan, 1996), smallholders often apply N fertilizer at too low rates 
and too late for obtaining good crop-yield responses (Heisey & Mwangi, 1996). 

Most smallholder farmers apply cattle manure — usually collected from 
enclosures (bomas, kraals) where cattle spend the night — but at rates too low to 
meet crop requirements and prevent decreases in SOM content. In the Heisey and 
Mwangi (1996) study, manures accounted for <10% of N inputs in Africa, or 
about 1 kg N ha" 1 yr" 1 . In intensively managed smallholder areas like the Kisii 
District of Kenya, applications of manure to the fields from cattle enclosures 
average 23 kg N ha" 1 yr" 1 , or about one-half of the total N inputs (Smaling, 1993). 
Manure is often diluted with soil when shoveled from cattle enclosures, and its 
quality and nutrient composition also is affected by the quality and quantity of 
fodder the animals eat (Murwira et al., 1995; Probert et al, 1995). The value of 
manure as a source of N ranges from high-quality manure that increases crop 
yields to low-quality manure that depresses crop yields due to N immobilization, 
with a critical threshold value of 1.25% N (12.5 g N kg" 1 ; Mugwira & 
Mukurumbira, 1986). 

The management of BNF is well established and practiced through the cul- 
tivation of N-fixing plants and the use of rhizobium inoculants (Giller & Wilson, 
1991; Giller et al., 1997, this publication). Trees provide N inputs in agroforestry 
systems by two processes, BNF and deep-nitrate capture. Grain legumes and 
herbaceous green manures supply N principally through BNF. Although the mag- 
nitude of BNF is methodologically difficult to quantify, overall estimates are in 
the order of 25 to 100 kg N ha" 1 per crop for grain legumes and as much as 280 
kg N ha- 1 yr- 1 for some herbaceous and woody perennials (Giller & Wilson, 
1991). BNF can supply considerable N inputs to crops via litter decomposition in 
soils, as long as these soils have enough available P. 

Legume cereal intercrops are widely grown as a means of reducing the risk 
of crop failure and providing households with improved diets. The potential N 
replenishment of a grain legume is a balance between fixed N and N in harvest- 
ed products (Giller & Cadisch, 1995). Unfortunately the potential for netN inputs 
by BNF with grain legumes is quite limited. Nitrogen fixation by peanut (Arachis 
hypogea L.) ranges from 68 to 206 kg N ha" 1 per crop, but most of it is removed 
at harvest (Giller & Wilson, 1991). Common bean is widely cultivated in East 
Africa, but it has such a low inherent capacity to fix N that it is likely to produce 
a negative N balance in the soil (Giller & Cadisch, 1995). Soybean [Glycine max 
(L.) Mem] has a high BNF capacity, but it concentrates N in the pods, adding lit- 
tle to the soil. Therefore, the contribution of BNF by commonly grown grain 
legumes with high N harvest index does not seem relevant to N replenishment. 

Short-term fallows of leguminous trees and herbaceous cover crops, how- 
ever, provide a practical means of N replenishment via BNF when grown in rota- 
tion with cereal crops. Two-year tree fallows of sesbania [Sesbania sesban (L.) 
Mem] or tephrosia (Tephrosia vogelii Hook, f.) have replenished soil N levels 
enough to grow three subsequent high-yielding maize crops in N-depleted, but P- 
sufficient soils in southern Africa (Kwesiga & Coe, 1994; Kwesiga et al., 1998). 
In Coastal West Africa, 6-mo herbaceous fallows of mucuna [Mucuna pruriens 
var. utilis (L.) DC] supply the N needs of one subsequent maize crop (Osei- 


Bonsu & Buckles, 1993; International Institute of Tropical Agriculture, 1995; 
Manyong et al., 1997; Galiba et al., 1997). In general, woody fallows accumulate 
larger N stocks than herbaceous ones because of their larger and continuing bio- 
mass accumulation (Szott et al., 1998). The residual effects of tree fallows are 
therefore longer than those of herbaceous fallows. 

There is evidence that non-N-fixing trees and shrubs of the genus Senna 
and Tithonia accumulate as much N in their leaves as N-fixing legumes, presum- 
ably because of their greater root volume and ability to scavenge nutrients from 
the soil (Szott etal., 1991; Garrity & Mercado, 1994, Gachengo, 1996). But it is 
important to note that these nonfixing trees are only cycling the N present in the 
soil, not adding inputs to the system, as happens via BNF in woody and herba- 
ceous leguminous fallows. Non-N-fixing trees and shrubs can only be considered 
to be N inputs when biomass is transferred from one field to another. 

Tree roots are often able to capture nutrients at depths beyond the reach of 
most crop roots. This can be considered an additional nutrient input in agro- 
forestry systems when such nutrients are transferred to the topsoil via the incor- 
poration and subsequent decomposition of tree litter. Hartemink et al. (1996) and 
Buresh and Tian (1997) detected subsoil nitrate levels in the order of 70 to 315 
kg N ha" 1 at 0.5- to 2-m depth in maize-based systems on Oxisols and Alfisols 
of western Kenya. They also found that sesbania fallows depleted this pool, thus 
capturing a resource that was unavailable to maize crops (Mekonnen et al., 1997; 
Fig. 1-6). The source of this nitrate pool is believed to be the result of the min- 
eralization of organic N in the topsoil, which is relatively high in these soils, fol- 
lowed by nitrate leaching into subsoil layers. The nitrate anions are then held in 
the oxidic subsoil by positively charged clay surfaces. 

Subsoil nitrate accumulation and its depletion was detected in East Africa 
decades ago (Mills, 1953; Kabaara, 1964), but such findings were not given prac- 
tical attention at that time. It is probable that trees also capture K at same depths 
in similar soils and thus help prevent K deficiencies. In order for nitrate anions to 
move, they must be accompanied by a cation; K is the main leachable cation in 
such soils. 

Nitrate accumulation in the subsoil is well documented in soils with sub- 
soils rich in Fe oxides that provide anion-exchange sites to hold nitrate ions 
(Kinjo & Pratt, 1971; Black & Waring, 1976; Cahn et al., 1992). Many such sub- 
soils, however, are highly Al-toxic, preventing significant plant root develop- 
ment, but subsoil acidity is not a widespread constraint in African soils cultivat- 
ed by smallholder farmers. 

The rotation of annual crops with short-duration fallows containing deep- 
rooted perennials holds promise as a way to use subsoil nitrate that would other- 
wise be unavailable to crops. This resource may not be replenished when crop- 
ping systems become more intense, as nitrate leaching from the topsoil may be 
diminished by more extensive crop root systems. The magnitude of captured sub- 
soil nitrate needs to be assessed in other soils, but soil chemistry indicates that 
subsoil nitrate accumulation will not be as significant in many other types of soils 
found in Africa. Nevertheless, there are 260 million ha of oxidic soils in Africa 
that have anion-exchange capacity in the subsoil. The use of this hitherto unrec- 


Nitrate N (kg ha'' 0.5 m layer' 1 ) 

ognized N source via its capture by deep-rooted trees is an exciting area of 
research in regions with oxidic, but not Al-toxic subsoils. 

Internal Flows 

Considering the inconsistent use of N fertilizers and the very limited returns 
of crop residues to the soil, most of the internal N cycling in maize-based small- 
holder systems in Africa results from the mineralization of soil organic N 
(Woomer et al., 1997). Such process may contribute most of the N required for 
low-yielding grain crops, until the labile soil organic N fractions (N capital) are 
depleted. Incorporating N-fixing trees and herbaceous legumes into the farming 
system enhances nutrient cycling and also provides the organic C and N neces- 
sary for maintaining N capital (Palm, 1995). 



A typical maize crop in smallholder African farms yields <lt ha" 1 of grain 
and requires a plant accumulation of <40 kg N ha" 1 ; a 41 ha" 1 maize crop requires 
100 kg N ha" 1 , and a 7 t ha" 1 maize crop requires 200 kg N ha" 1 (Sanchez, 1976). 
Approximately two-thirds of this N is accumulated in the grain and will be 
exported during harvest. Much of the remaining N is located in the stover and will 
not necessarily be cycled back to the soil because crop residues are frequently 
burnt, sold, or fed to livestock, and the manure produced is applied to higher- 
value crops growing in other fields. Loss processes such as soil erosion, leaching, 
and denitrification represent N outputs. In sandy soils of the Sahel, 26 to 47% of 
the surface-applied urea is lost through ammonia volatilization (Christianson & 
Vlek, 1991). 

Can Nitrogen Demands be Met Biologically? 

Herbaceous leguminous cover crops provide sufficient N inputs through 
BNF to meet the needs of one subsequent maize crop. In Africa, the main species 
used are of the genus Mucuna, Crotolaria, Pueraria, Dolichos, and Desmodium 
(Balasubramanian & Blaise, 1993; Wortmann et al, 1994). The use of mucuna in 
short-term fallows interplanted with maize or planted during the dry season is 
expanding rapidly in Benin and Ghana (Osei-Bonsu & Buckles, 1993; Manyong 
et al., 1997). Mucuna fixes N and smothers imperata [Imperata cylindrica (L.) 
Rausch.] (International Institute of Tropical Agriculture, 1995). Herbaceous legu- 
minous fallows because of their shorter duration and lower biomass accumulation 
provide lower N inputs than woody leguminous fallows (Szott et al., 1998). 

Improved herbaceous and woody leguminous fallows, therefore, provide 
excellent options for managing N biologically, provided that the soil is sufficient 
in P and that farmers are willing to make land available for crop-fallow rotations. 
Empirical evidence in Benin, Ghana, and Zambia show that thousands of farmers 
are taking advantage of such an option (Galiba et al., 1997; Kwesiga et al., 1998). 

At high-crop-yield levels comparable to those of commercial farms in the 
temperate zone (above 6 tha" 1 of maize grain), organic N inputs are likely to be 
insufficient and therefore must be supplemented with inorganic fertilizers. The 
senior author has seen some farmers in Chipata, Zambia doing this. They rely on 
2-yr sesbania or tephrosia fallows to provide the basal N and then use N fertiliz- 
ers at top dressing time if the rainfall is favorable and the crop shows N defi- 
ciency. If the expected production is not promising, farmers may save the N fer- 
tilizer for the next year. This is an excellent example of the strategic use of fer- 
tilizers (Sanchez, 1994). 

Replenishing Nitrogen Capital 

Organic inputs have an important advantage over inorganic fertilizers with 
regard to fertility replenishment — they provide a source of C for microbial use 


(Fig. 1-5). According to Palm (1995), the recovery by the crop of N from the 
leaves of leguminous plants incorporated into the soil (10-30%) is generally 
lower than the recovery from N fertilizers (20-50%). Much of the remaining 70 
to 90% of the applied organic N not used by crops or leached is incorporated into 
labile pools of soil organic N and C. Soil microorganisms require C substrate for 
growth and to use the N from organic inputs to form soil N capital. Part of the N 
bound in the more recalcitrant fractions in the organic inputs also will increase 
soil organic N (Giller et al., 1997, this publication). Inorganic fertilizers do not 
contain such C sources, and therefore much of the fertilizer N not used by crops 
is subject to leaching and denitrification losses in the absence of crop residue 

Long-term experiments in Africa provide indirect evidence in support of 
the combined organic and inorganic approach to replenishing N and C capital. 
Kapkiyai (1996) reports a 29% loss of total soil N (1.06t N ha" 1 in the top 15 cm) 
when maize and beans were grown in rotation for 18 yr without nutrient inputs 
and with crop residues removed in Kabete, Kenya. The same loss took place in 
plots with the recommended fertilizer applications but no residues returned; how- 
ever, when fertilizers and manures were added and the maize stover was retained, 
the decline in total topsoil N was reduced by one-half. Organic inputs or the recy- 
cling of crop residues apparently provided the soluble C necessary to reduce N 
depletion in this fertile soil. 

In sandy soils of the Sahel, Pieri (1989) also found that additions of N fer- 
tilizer alone did not increase soil C or N stocks. But fertilizers plus organic inputs 
(crop-residue returns, manures, and composts) increased soil N and C stocks, 
except in extremely sandy soils where there are too few clay particles to protect 
newly formed SOM from decomposition. 

There are some extreme situations in Africa where virtually all organic 
resources are depleted or are used to meet more pressing needs. Much of the 
Central Ethiopian highlands have been converted from forest to smallholder agri- 
cultural landscapes essentially devoid of trees. Crop residues are fed to livestock, 
while manure and even roots are used as cooking fuel. Insecure land tenure dis- 
courages the replanting of trees, thus exacerbating the problem. To break clear of 
this vicious cycle, N fertilizers must be applied along with whatever organic 
inputs become available from planting trees. 

Research to date has mainly compared inorganic vs. organic sources of N 
with little consideration of the nutrient content of organic sources. The quantita- 
tive interaction between organic and inorganic sources of N is essentially a new 
subject of research in the tropics (Palm et al., 1997b, this publication). 

Thejoint organic-inorganic N replenishment strategyjust described is not 
new; it has been used for centuries in temperate agricultural systems, where N 
fertilizers together with crop rotations, winter cover crops, manure applications, 
and the full incorporation of crop residues provide sufficient N and C inputs to 
gradually increase soil N and C stocks (Buol & Stokes, 1997, this publication). 
What is new is the potential to do something similar to replenish soil N in the 
tropics with systems that add N inputs in situ and are consistent with the con- 
straints of smallholder farmers. 



Soil-fertility replenishment requires a set of accompanying technologies 
and policies to be effective in raising and sustaining food production. By itself, 
soil-fertility replenishment is a necessary but not sufficient condition for increas- 
ing per capita food production in Africa. 

Soil Conservation 

First and foremost, soil erosion control technologies must be present in 
order to keep the nutrient capital investment in place and to prevent nutrient pol- 
lution of rivers, lakes, and groundwater. Where the labor to land ratio is high, 
such as in parts of the East African highlands, various labor-intensive soil con- 
servation technologies are financially attractive and widely used (Cleaver & 
Schreiber, 1994). A P investment program that does not include contour hedges 
or other erosion control technologies is likely to do more harm than good. 
Fortunately, there are well -proven biological methods of erosion control, such as 
growing leguminous hedges or vegetative filter strips along the contours (Kiepe 
& Rao, 1994; Garrity, 1996). Soil conservation technologies are more readily 
adopted if they provide useful by-products. The technologies mentioned also can 
provide fodder, fuelwood, fruit, and biomass for transfer to adjacent fields. 
Farmers' willingness to undertake soil conservation measures will increase with 
population density and with policy reforms that make intensive farming more 
profitable (Cleaver & Schreiber, 1994). 

Sound Agronomic Practices 

There are several positive feedbacks to soil-fertility replenishment. It is 
more likely that sound agronomic practices will be profitable in replenished areas 
than in depleted ones. Such practices include the use of improved crop 
germplasm, integrated pest management, crop-residue returns, crop rotations, 
supplemental irrigation, and maintenance fertilization. Many of these practices 
are not worthwhile in nutrient-depleted soils because low crop yields give nega- 
tive returns to such investments. 

Fertility replenishment also enables farmers to intensify and diversify their 
production. They may shift from growing low-value crops to growing vegetables, 
livestock, or trees that produce high-value products, which may add economic 
sustainability through product and income diversification (Cleaver & Schreiber, 
1994; Tomich et al., 1995; Sanchez & Leakey, 1997). Such diversification also 
will contribute to environmental resilience through increased plant biodiversity. 

Some pest problems related to low soil fertility are diminished when fertil- 
ity is replenished. For example, in addition to replenishing N, 1-yr fallows of ses- 
bania have been found to encourage suicidal germination of the parasitic weed 
striga [Striga hermonthica (Del.) Benth.], a major maize pest in western Kenya, 
reducing its seed pool by one-half (Oswald et al., 1996; Amadou Niang, 1997, 
personal communication). 


Some negative feedbacks also are likely to occur. New pest problems may 
arise as a result of increased soil fertility because of higher plant biomass and 
moisture in the fields. Deficiencies of other nutrients, such as K, may become evi- 
dent due to the higher nutrient offtake by high-yielding crops. In such cases, how- 
ever, the added fertilizer requirement should not be considered a capital invest- 
ment but a recurring cost of production that should be paid for by increasing crop 


Specific districts or other divisions of a country that are affected by severe 
nutrient depletion can be identified for nutrient replenishment projects. Project 
development is best designed and conducted with farmers and communities to 
assure their involvement from the beginning (Ashby, 1986, 1987). 

Case studies are useful in bringing a more direct perspective. We describe 
three recent case studies in biophysically and socioeconomically contrasting 
areas of Africa: (i) western Kenya, a densely populated, high-potential area with 
N and P depletion, high P-sorbing soils, and secure land and tree tenure, (ii) east- 
ern Zambia, a less densely populated, medium potential area with soils depleted 
only of N and secure customary land and tree tenure, and (iii) Central Burkina 
Faso, the least densely populated area with low potential, low initial levels of N 
and P, insecure land tenure, and no tree tenure. 

Western Kenya Case Study 


The highlands of western Kenya, which are part of the Lake Victoria Basin, 
have one of the densest rural populations in the world — 500 to 1200 people km" 2 
(Hoekstra & Corbett, 1995). Annual rainfall ranges from 1200 to 1800 mm with 
a bimodal distribution. Elevation averages 1200 m, and the main soils are high P- 
sorbing Alfisols and Oxisols, originally quite fertile but now widely depleted of 
N and P. Characterization studies identified declining soil fertility as the main 
factor limiting crop production (Hoekstra, 1988). 

There are about 6 million people and 2 million farms in a total area of 10 
000 km 2 , with an average farm size of 0.5 ha of which about one-third is planted 
to maize. With maize yields often as low as 1 t ha" 1 over two seasons and house- 
holds needing >1000 kg yr ' of maize for food security, most households are only 
producing enough maize to feed themselves for a few months. They must pur- 
chase maize on the market during the remaining months or endure hunger peri- 

About 80% of farms in Vihiga, Siaya, Busia, and Kisumu Districts are 
severely deficient in P (<5 mg bicarbonate-extractable P kg" 1 soil), and most are 
deficient in N when P deficiency is overcome (Shepherd & Soule, 1998; Bashir 
Jama, 1997, personal communication). Heavy striga infestations frequently occur 
in N-depleted soils. The irony is that the main soils of the region (Rhodudalfs and 


Table i - 1 . Basket of choices to replenish soil fertility in small farms of western Kenya. 

Interplant an improved fallow of Sesbania. tephrosia. or crotolaria (Cn/lolarin gniluimiumi Wight & 
An). I during the long rains into the maize crop and after the second weeding. Let the fallow 

ill I file fall II i in 111 iii ii ' I 

about 150 kg N ha" 1 to the soil. 
{ 'onstrncl contour bunds, ditches, or terraces and plant tithonia along them after the maize harvest 

i n m c i oi trol in inn i i '1 ni hi I field 
Planl additional tithonia in hedges along farm and field boundaries or in vacant land to accumulate 
about 0.7 t of fresh biomass every 6 mo per farm (aboul 500 m of 1-m-wide hedges). 
Transfer tithonia biomass to the fields (at the rate of 2 t dry material ha" 1 ) before maize 
planting. Tithonia adds about 60 kg N, 6 kg P, and 60 kg K ha" 1 to the soil and increases the 
availability of P fertilizers. 
pply ot i i i 1 1 ii i ii nanure prod I on farm 

Apply a reacth e [ h h I it I ivei replenishment nil t kg I I i 

pel farm, and incorporate it with tithonia and the leaf materials from leguminous fallows. 
Mi In r i i 1 i i i i i quent crops. 

Shift some of the replenished fields from maize to high-value vegetables such as sukuma wiki 
/ in lly agroforesti i dti hi h 

value fruits, pharmaceuticals or high-grade timber. 

Eutrudoxs) are considered among the most productive soils of the tropics 
(Sanchez, 1976). 

Farmers realize the value of fertilizers in western Kenya. About 40% of 
them use some DAP, but at lower than recommended rates and often too late for 
optimum timing of applications (Swinkels et al., 1997). In spite of the extreme 
land pressure, about 52% of the fanners leave a portion of their farm in weedy 
fallows (Swinkels et al, 1997). Fallowing is often not a matter of choice, because 
either the land is severely depleted or labor and agricultural inputs are not avail- 
able (Amadou Niang, 1997, personal communication). This provides an entry 
point for the organic inputs to be grown in situ. 


Five years of collaborative research have shown that the use of organic and 
inorganic inputs can replenish soil fertility and decrease striga infestations to 
manageable levels. In addition to researcher-managed on-farm trials, there are 
currently a total of about 1000 farmers experimenting with the improved tech- 
nologies through district extension services, NGOs, and a team of national and 
international centers working there as part of the African Highlands Initiative 
(Wang'ati & Kebaara, 1993; ICRAF, 1996). About one-half of the farmers are 
experimenting with improved fallow technologies, the rest with biomass transfer 
technologies, and a few with P fertilizers. Okalebo and Woomer (1996) suggest 
the target should be to replace P lost over 20 yr (an average of =250 kg P ha" 1 ) in 
farmers' fields where P is the most limiting nutrient. Here, the use of PR appears 
promising due to its sub-regional availability, suitable soil pH, and price advan- 
tage. The replenishment strategy is based on the steps shown in Table 1-1. Some 
research results have been shown in Fig. 1-4. 

Impact Models 

The first step in analyzing the feasibility of a replenishment strategy is 
determining if it is profitable at the farm level. An economic-ecological farm 


model for assessing the impact of soil-fertility replenishment on farm income and 
nutrient balances has been developed for the East African highlands (Shepherd & 
Soule, 1998). This or similar models can indicate the potential profitability and 
adoptability of new practices as well as their ecological consequences. 

The model simulates the impact of current and improved agroforestry and 
other soil-management practices at the farm scale on nutrient availability, nutri- 
ent cycling and losses, and plant and livestock production. It also provides vari- 
ous measures of financial returns. Farm size, crop allocation, and soil-manage- 
ment practices are based on typical values for farmers with different levels of 
resource endowment. 

Simulations have been carried out to analyze the farm-level impact of P and 
N replenishment for farmers with low- and high-resource endowments in the 
densely populated Vihiga District of western Kenya. Soils in the district are gen- 
erally high P-fixing and P-deficient. A typical low-resource farmer has about 0.2 
ha of land and grows mainly maize and beans. About 55% of the farmers in the 
district have been classified in the low-resource endowment category, and 35% of 
the farmers are classified in the medium-resource endowment category. Both cat- 
egories have per capita incomes of <1 U.S. dollar per day (ICRAF 1996; Narayan 
& Nyamwaya, 1995). High-resource endowment farmers make up only about 
10% of the farming population. They typically have about 1.6 ha and farm more 
intensively, incorporating dairy cows (Bos taunts) and significant amounts of 
purchased inputs (Shepherd & Soule, 1998). 

Simulations over 20 yr were made of the existing system and three 
improved systems based on a one-time investment dressing of Minjingu PR (250 
kg P ha" 1 ) with three alternative sources of N: (i) urea at the rate of 60 kg N ha" 1 
yr" 1 , (ii) transfer of green biomass oftithonia from existing farm borders (1.9 t 
dry green biomass ha" 1 yr" 1 , producing 60 kg N ha" 1 ), and (iii) improved fallow 
of sesbania grown on one-half of the field area and rotated with the maize and 
bean crops. 

For the low-resource endowment farmers, the simulation results demon- 
strate large potential increases in crop yields from P replenishment (Fig. l-7a), 
whichever source of N is used. Nitrate leaching (Fig. 1 — 7b) is initially highest in 
the existing system (no inputs) because the N mineralized from organic matter 
cannot be fully used by the P-limited crops, butN leaching in this system declines 
as SOM levels decline. When P is applied nitrate leaching is reduced, especially 
in the tree-based improved fallow system (IF). Soil organic P (Fig. l-7c) decreas- 
es under the existing system, but it is maintained or increased with P replenish- 


Net farm income (the value of crop output less the cash costs of production) 
increased by 80 to 160% over the existing system by Year 2 of the simulation 
(Fig. l-7d); however, farm returns, which is net farm income less family labor 
valued at the market wage, is lower with biomass transfer (BT) than with urea or 
IF due to the large labor requirement of BT. Farm returns will vary with the actu- 
al opportunity cost of labor, which is probably below the market wage (U.S. dol- 





s « 

1 1" 

-»-PR+BT -»-PR+iF 

5 10 tS 20 

Fig. I i hi I i n ir altennit [in terns ti typical i i i i t u li i n n Ken nutrient input liospl rock (PR) at 250 

kg P ha"' plus urea at 60 kg N ha- 1 , PR plus biomass transfer (BT) from farm boundaries, and PR plus improved fallow I IF) with sesbania. P yield is P in harvested 
products (maize and bean grain and fuel wood). Soil N leached is at 2-m soil depth. Net income is based on 1 US$ = 55 Kenya shillings. 


Table 1-2 Simulated net present values (NPVi based on a discount rate of20% ai 

I'oui lternati\ land-u tem lor a typical ui | ir farm (0 ha) in \ estein Kenya. The 

four land-use systems are the existing system, which is based on maize-bean intercropping and use 
no i I nutrient inj til id t impi i m h I n n m nt d it i (I 

kg P ha" 1 as phosphate rod, (PR) and then either urea at 60 kg N ha" 1 yr "', transfer ofbiomass of 
tithonia (1.9 t dry biomass ha" 1 ) from farm boundaries, or an improved fallow of sesbania on one- 
half of the field area each year. 1 US$ = 55 Kenya shillings. 

Existing PR + PR + biomass PR + improved 

NPV of net farm income 
NPVoffarm returns 

Investment in P (first year only) 
Annual N investment 

lars 0.90 d" ) for the low- and medium-resource farmers but close to the market 
wage for the high-resource category. 

At a discount rate of 20%, the NPV of net farm income and farm returns 
are much higher for all three replenishment scenarios than for the existing system 
(Table 1-2), suggesting that P replenishment is indeed profitable for low-resource 
farmers with P-deficient and high P-sorbing soils. Whether an individual farmer 
might prefer to combine PR with urea, improved fallow, or biomass transfer 
depends on the cash and labor constraints faced by that farmer; however, the ini- 
tial investment required for P replenishment (U.S. dollars 40 per 0.3 ha per farm) 
is greater than the annual farm income generated by the existing system and about 
10% of the average annual household income of about U.S. dollar 420 for a low- 
resource farmer (Shepherd & Soule, 1998). Such households are unlikely to be 
able to undertake an investment in PR without some form of financial assistance. 

The high-resource endowment farmers, on the other hand, do not profit 
from P replenishment since their current practices already include annual P inputs 
and large manure applications from their dairy enterprise. Years of good land hus- 
bandry have maintained soils at productive levels. The high -resource endowment 
farmers are able to farm profitably and sustainably because they are not as severe- 
ly affected by capital, labor, and information constraints as are small, low- 
resource farmers. For high-resource farmers, the capital constraint has been 
released by access to significant amounts of off-farm income. Indeed, high- 
resource endowment farmers are partly defined by their access to off-farm 
income and the use of such income for on-farm investments (Crowley et al., 

Eastern Zambia Case Study 


In less densely populated areas (20 to 40 people km"') of the eastern 
province of Zambia, an area typical of the Miombo woodlands of southern Africa 
(White, 1983), grass fallows of 1 to 5 yr coexist with maize cultivation, which in 
a sense is a form of shifting cultivation in tropical savannas. A diagnostic and 


design survey in Katete and Chipata Districts (Ngugi et al., 1988) revealed a seri- 
ous breakdown of traditional strategies to sustain production of food, fodder, and 
fuelwood. Declining N fertility was identified as the major problem responsible 
for low yields of the main staple food crop, maize. Phosphorus is not yet an 
important constraint in this region. Maize no longer receives N fertilizers since 
the removal of subsidies, which tripled or quadrupled the cost of fertilizer N rel- 
ative to the prize of maize (Heisey & Mwangi, 1996). There is a hunger period 
when the maize supplies run out before the next harvest. Given the relatively 
large farm size (3 to 5 ha) and the widespread use of grass fallows, improved fal- 
lows seemed the logical entry point in this area. 


The strategy developed was to use leguminous fallows to accumulate N 
from BNF, smother weeds, and improve soil physical properties (Kwesiga & 
Chisumpa, 1992). The main species identified were sesbania, tephrosia, and 
pigeonpea. Two-year-old sesbania or tephrosia fallows doubled maize yields dur- 
ing a 6-yr period, in comparison with continuous unfertilized maize production 
(Kwesiga & Coe, 1994; Kwesiga et al, 1998). This was accomplished in spite of 
2 yr without crop production while the fallows were growing. 

Research also is in progress on alternative sesbania cultivars and on other 
species, cheaper establishment methods such as bare-root seedlings, and combin- 
ing improved fallows with top dressings of N fertilizers to push yields to a high- 
er plateau (Kwesiga et al., 1998). Researchers also are examining farmer percep- 
tions, local policies to protect the fallows from grazing, and the overall adoption 
potential (Franzel, 1998). 

On-farm research started by establishing solid relationships with the exten- 
sion staff, and through them to the farmers. The approach was initially to select a 
village near a farmer training center, which would later be used for demonstra- 
tions. Meetings were then arranged where both the researchers and extension staff 

Mwinxa Jera Nthani 

x yields in five farmer-designed and 


interacted with farmers and discussed the causes of low maize yields, the farm- 
ers' fallowing practices, and the potential of improved fallows. Such visits gen- 
erated much discussion among farmers and confirmed that the farmers were gen- 
uinely interested in the technology. The research-extension team increased the 
frequency of field days so that they coincided with the major phases of improved 
fallows: the nursery, the fallow, and crop (Kwesiga et al., 1998). 

About 158 farmers initiated researcher-designed, farmer-managed trials, 
each with 400 m 2 plots of improved fallows. These farmers represented a range 
of high and low income, male and female, and oxen and hoe users. In the trials, 
farmers selected one of six options of improved fallow technologies. More than 
70% planned to continue using the technology during the next season. 

In farmer-designed and -managed trials, fanners were provided seed or 
seedlings and advice on options available, such as fallow length, tree density, and 
planting method. They were left to design their own trials, planting trees where 
they wished in their own farms. The main purpose of this type of trial is to under- 
stand how improved fallows are accepted by farmers into their existing farm 
practices. The number of farmer-designed trials increased from five in 1993, to 
37 in 1994, and to 797 in 1995. Many farmers who initially started off as 
researcher-designed, farmer-managed trials also planted farmer-designed trials 
after experiencing the benefits of improved fallows. Usually, they produced 
planting materials in their own farms. 

During 1996, five farmers harvested maize following 2- or 3-yr sesbania 
fallows. Maize yields were impressive (Fig. 1-8), and yields for four of the five 
farmers were comparable to those achieved from fully fertilized controls. These 
results, showing that sesbania fallows increased maize yields without fertilizers, 
triggered enthusiastic responses from a large number of farmers, extension staff, 
NGOs, and development agencies. The challenge now is to increase the technol- 
ogy adoption from 4000 farmers now practicing it (Kwesiga et al., 1998) to mil- 
lions of farmers in southern Africa. 


The rotation of a 2-yr sesbania fallow followed by 3 yr of maize produced 
92% more wealth (NPV of US$ 588 ha" 1 ) than the current situation, 5 yr of con- 
tinuous maize cultivation without N inputs (Table 1-3). The most profitable 
option, however, was the recommended rate ofN fertilizer (112 kg N ha- 1 per 
crop) — an option most farmers are no longer able to consider. 

Table 1-3. Cumulative discounted net benefits for 2-yr sesbania improved fallows followed by 4 yr 
nonfertilized maize as compared with continuous maize cropping at Chipata. Zambia, with a dis- 
count rate of 20% (Kwesiga et al., 1998). 

Continuous maize, no fertilizer 

2-yr sesbania fallow-nonfertilized maize 

Continuous maize (1 12 kg N ha- 1 yr ' ) 


The returns to labor may be more important to farmers than the returns to 
land in this area, where labor is more scarce than land. The returns to labor from 
a 2-yr fallow, 3-yr maize rotation were US$ 3.45 per day, which is 70% more than 
from continuous monocropped maize without fertilization. If farmers plant 2 ha 
of maize after 2-yr fallows every year, they would overcome food insecurity and 
achieve a maize surplus of 205%, except in drought years. 

Numerous sensitivity analyses were undertaken (ICRAF, 1995). They 
included changes in the wage rate, cost of seedlings, maize yields, and fuelwood 
prices and an investigation into how changing occurrences of drought affected 
fallow performance. In virtually all reasonable scenarios, the 2-yr fallow was 
shown to be more attractive than the existing practice except for one with an 
extremely high discount rate (over 40%). 

Central Burkina Faso Case Study 


Unlike the previous two, this case study focuses on strategies formulated at 
the national level, drawn largely from a series of technical and policy proposals 
(Mokwunye et al., 1996; De Jager & Smaling, 1996; Bikienga, 1997; Dembele, 
1996). The Central or Mossi Plateau of Burkina Faso is a semiarid, low-potential 
area typical of the Sudanian belt of Sahelian West Africa. It is characterized by 
low (600-800 mm yr" 1 ) and highly variable rainfall, naturally infertile sandy 
Alfisols very low in N, P, and SOM, high soil erosion risk, and cropping systems 
based on low yields of sorghum [Sorghum bicolor (L.) Moench] and pearl millet 
[Pennisetum glaucum (L.) R. Br.] in between scattered parkland trees. There is 
low (20 people km"") but increasing population density, unavailability of fertiliz- 
ers to smallholders, and insecure land ownership by individuals, while tree tenure 
is held by the government (Sanders et al, 1996). There is a relatively large PR 
deposit (Kodjari) under production for local use. About 10% of the country's land 
under cultivation has stone lines, diguettes, or other soil erosion control devices. 
Fertilizer use is largely limited to cotton (Gossypium hirsutum L.) growing areas 
run by parastatals (Bikienga, 1997). 

Extensive on-station and on-farm research shows that P deficiency must be 
overcome before N responses can be realized. The reactivity of the Kodjari PR is 
low. It is not recommended for direct application, but it could be used with par- 
tial acidulation (Bationo & Mokwunye, 1991). Even with the low pH values of 
most soils (4.5 to 4.8), 2 to 3 yr are required to achieve full crop responses to this 
PR. Solubilizing Kodjari PR with organic acids derived from composts, however 
appears promising (Lompo, 1993). 

Action Plan 

The Burkina Faso government eliminated fertilizer subsidies in 1987. In 
1995, a multidisciplinary Soil Fertility Management Unit (Unite da Gestion pour 
la Fertilite des Sols) was established at the cabinet level to design strategies and 
action plans to replenish soil fertility in the most depleted areas with a combina- 
tion of fertilizers, improved germplasm, and the development of inputs and out- 


puts markets (Dembele, 1996). Nationwide sensitization workshops enabled the 
Soil Fertility Management Unit to increase awareness of soil-fertility depletion as 
well as to combine efforts to improve the availability of nutrient inputs along with 
the marketing of farm products. The basic strategy is to use Kodjari PR to correct 
P deficiency as a one-time capital investment paid by the government via donors. 
This calls for a basal application of 400 to 600 kg ha- 1 of Kodjari PR (12% P), 
which is equivalent to 48 to 72 kg P ha" 1 , supplemented by annual maintenance 
PR applications of 12 to 24 kg P ha" 1 and urea (Bikienga, 1997). 

Market research shows that the private sector is reluctant to sell fertilizers 
on credit to smallholder cereal producers due to high investment risk from 
drought (Dembele, 1996). Therefore, the government feels it must take over input 
distribution. The plan also includes improving the technical knowledge about fer- 
tilization by private distributors and increasing the role farmer organizations play 
in fertilizer and cereal trade. 

A full country-wide replenishment, including maintenance doses, requires 
4.3 million t of Kodjari PR ore annually. The current estimate is that Kodjari PR 
deposits can supply that rate for 60 years (Bikienga, 1997). Furthermore, priori- 
ty is to be given to the 700 000 farmers that have invested in soil erosion control 
and compost pits, and those that grow grain legumes in rotation with cereal crops. 
The plan is to be implemented by farmer associations that focus on women. 
Bikienga (1997) calculates the cost of Kodjari PR at farm gate to be US$ 150 f' 
similar to the cost of Minjingu PR in western Kenya. Total development costs for 
5 yr, including refurbishing the current Kodjari PR production site, purchasing 
90 000 t of PR, 3000 village demonstrations, training, and monitoring are esti- 
mated to be about US$ 25 million (calculated from Bikienga, 1997). 

There are no specific provisions in this plan for organic inputs, except for 
growing grain legume crops in rotation, which as discussed earlier is not a real- 
istic way to replenishing N. Phospho-composts appear to be a more feasible 
option (Lompo, 1993), but their costs need to be fully assessed. Other options that 
may be considered are leguminous tree fallows capable of growing during the dry 
season, like in Zambia and Kenya, or biomass transfers like in Kenya. Both 
options need to be researched. The woody fallow species must be adapted to the 
climate and soil stresses of the Sahel. Research near Bamako, Mali indicates that 
gliricidia [Gliricidia sepium (Jacq.) Walp.] fodder banks produce significant leaf 
biomass, about 3 tha"' yr" 1 of dry matter (ICRAF, 1995), with a potential accu- 
mulation of 60 to 90 kg N ha- 1 yr 1 . 


A World Bank-sponsored PR feasibility study based on data from the cen- 
tral and northern regions of Burkina Faso shows aNPV ofUSS 396 ha- 1 for the 
local PR, using a 20% discount rate for private benefits, 10% for national bene- 
fits, and 2% for global benefits (World Bank, 1996, unpublished study). At low 
rates of application, imported soluble P fertilizers are more profitable than 
Kodjari PR. In response to this finding a better option might be to use the low- 
5 PR as the basal application and lower rates of superphosphate as yearly 
plications. It is relevant to note that this study did not include the use of organ- 


ic inputs. This action plan is still in its formative stage and unlike the previous 
one, no pilot trials have been conducted at the time of this writing. 


The three different case studies show that various fertility replenishment 
strategies are profitable. Despite this, smallholder farmers face daunting con- 
straints to fertilizer use (Runge-Metzger, 1995). In the long-run, the alleviation of 
these constraints through improved government policies will provide farmers 
more incentives to undertake soil replenishment investment. 

The soil-fertility replenishment approach should emphasize the provision 
of truly public goods such as infrastructure and improved technologies, the 
removal of distorting incentives in agricultural input and output markets, and the 
encouragement of rural credit markets. The provision of credit at reasonable 
interest rates is viewed as particularly critical because investment in nutrient 
replenishment will be profitable in most cases, despite the poor marketing condi- 
tions of many rural areas. 

These policies may not, however, address the urgent short-term needs of 
resource-poor farmers. One short-run strategy might include the improvement of 
the availability and delivery of P fertilizers and both organic and inorganic N 
sources to farmers. This may take the form of facilitating importers in fertilizer 
acquisition, providing storage facilities, provision of germplasm for fallows and 
biomass transfer, and extending credit to wholesalers or retailers. This also 
should include better training of extension agents on adapting blanket nutrient 
input recommendations to specific farmer needs and nursery development for 
providing germplasm of the organic sources; however, improving fertilizer sup- 
ply is not sufficient to encourage the uptake of P by farmers as demand is influ- 
enced by price and capital. 

A short-term replenishment strategy must therefore also consider either the 
options of increasing credit to farmers or cost-sharing. If capital is truly the major 
constraint and P replenishment has strong residual effects, then one presumes that 
farmers would have sufficient incentives and profits with which to reinvest in 
maintenance nutrient inputs. In this case, the level of government intervention 
could be phased out, especially as the longer-term enabling policy changes begin 
to take effect. Even in the long-run, however, there can be some scope for direct 
government intervention in P fertilizer and germplasm delivery if social returns 
from replenishment significantly exceed private returns to fanners under liberal- 
ized market conditions. In this case, the level of farmer investment in P fertiliz- 
ers and tree nurseries will be less than optimal from society's point of view. 

Credit for the Poor 

Local access to credit at reasonable interest rates will be needed to finance 
costs that do not fall in the category of natural capital investments (improved seed 
of food crops and high-value plants, integrated pest management, N fertilizers, 
; fertilization of other nutrients). Mohammed Yunus, a banker, 


created the Grameen Bank in Bangladesh based on the explicit goal of alleviat- 
ing poverty. It gives small loans at market interest rates. In the process Yunus 
broke several banking rules: (i) the less money a person has, the more credit wor- 
thy that person and (ii) wage employment is less important than the possibility of 
becoming self employed. The best loan performers are rural women, which con- 
stitute 94% of the 2.1 million Grameen Bank's borrowers. The Bank has a loan 
recovery rate of more than 90%; it is now known as the Bank of the Formerly 
Poor (M. Yunus, speech at the World Food Prize Symposium, Des Moines, Iowa, 
18 October 1996). 

Peer pressure is a key component of the Grameen Bank's success. Out of a 
group of borrowers, the second two borrowers receive their loans only after the 
first two repay their loans, and the group leader is usually last in the group to 
receive a loan (Gladwin et al., 1997, this publication). Microcredit facilities, fol- 
lowing the Grameen Bank model are now being used in Sasakawa Global 2000 
projects in Ethiopia and Benin for loans up to 50% of the cost of inputs, with 
recovery rates of 90% or higher (Quinones & Takele Gebre, 1996). This is cer- 
tainly a model to follow in order to finance N fertilizers, hybrid seed, and other 
recurring costs for recapitalized soils. Furthermore, a recapitalized soil is likely 
to reduce the riskiness of a farmer in the eyes of a creditor. 


Based on the concept of the different kinds of capital described by 
Serageldin (1995), we concluded that soil fertility is most definitely a form of nat- 
ural capital. Benefits to farmers will be large. As such, they should be expected 
to contribute land, labor, and even capital towards the replenishment of their 
soils. Capital constraints, however, may prohibit significant farmer investment in 
P. Government intervention to promote credit is called for. If societal benefits of 
a regional P-replenishment project exceed the costs, cost-sharing mechanisms 
may be warranted, 

Izac (1997, this publication) identifies organic inputs and PR as two types 
of natural capital that can be considered as capital investments, because they gen- 
erate international public goods out of positive environmental externalities. 
Capital investments are different from subsidies in that they have a profit expec- 
tation in the long-term — an explicit return on the investment — while subsidies are 
short-term removals of constraints. Therefore, it may be advisable for society to 
assume some of the costs involved in moving farmers from unsustainable to sus- 
tainable production systems, in recognition of the socially and environmentally 
desirable externalities involved (Cleaver & Schreiber, 1994; Izac, 1997, this pub- 

In order for soil-fertility investments to have national or global benefits, 
they must be adopted at a large scale. Soil-fertility investments in only individual 
scattered farms will not provide national or global benefits. Action plans, such as 
the one for Burkina Faso (Bikienga, 1997) start gradually but are clearly aimed 
at recapitalizing large parts of the countryside. 

We, therefore, suggest that national and global societies invest in actions 
that increase the soil's nutrient capital in the long term. This means investing in 


large, one-time corrective applications of P fertilizers and in organic inputs that 
build up N, P, and C in the SOM (Izac, 1997, this publication). It also means not 
investing in N fertilizers as long as they do not build soil nutrient capital under 
present smallholder conditions, which do not include accompanying organic 

The efficiency argument against subsidies (see Gladwin et al., 1997, this 
publication) is made irrelevant by the cost-sharingjust described, when econom- 
ic and environmental externalities are taken into account. Izac (1997, this publi- 
cation) describes several long-standing cost sharing schemes between farmers 
and governments in developed countries, including assistance to farmers for 
planting trees to meet soil conservation objectives in the USA. Certainly some- 
thing similar could be applied in tropical developing c 


Combinations ofP fertilizers and organic inputs can replenish soil nutrient 
stocks in Africa and restore service flows approaching their original levels. Such 
restoration is in essence a long-term investment in the rebuilding of a country's 
stock of natural capital. We believe that the way forward is a cost-shared initial 
capital investment to purchase P fertilizer and germplasm to grow organic inputs 
combined with effective micro-credit for recurring costs such as V fertilizers and 
hybrid seed. 

Given the positive social and environmental externalities associated with 
soil-fertility replenishment, an equitable cost-sharing mechanism can be devel- 
oped and implemented, similar to existing ones in countries belonging to the 
Organization for Economic Co-operation and Development (OECD) that deal 
with positive environmental externalities (Izac, 1997, this publication). Cost shar- 
ing of the capital investments should be done on the basis of whoever benefits 
should pay. Subsidies, in our view, have a limited role to play. Being a one-time 
only investment, some of the dark sides of subsidies such as continuity and 
dependency are not likely to play a major role. 

Phosphate resources are certainly abundant in Africa. There is no question 
as to the essential role of P inputs regardless of whether P is replenished via direct 
applications of PR, combinations of PR with organic inputs to help its solubility, 
or via superphosphates. 

One of the main arguments against the use of organic N inputs is their low 
N concentration, in comparison with inorganic N fertilizers; however, when a 
leguminous fallow grown during the dry season or when no crops are in the field, 
can accumulate 100 to 200 kg N ha" 1 though BNF or deep nitrate capture, the low 
concentration argument becomes irrelevant. Although this in situ N production is 
not free, as it requires labor inputs, it certainly does not require significant cash 
inputs or entail transport costs. The biological approach to N replenishment must 
be viewed from a different perspective, now that short-term improved legumi- 
nous fallows are becoming a reality in Africa. 

A soil-fertility strategy for Africa must effectively address well-known 
n a novel way. The international donor community is asking much of 


smallholder African farmers: Borrow money at 20% interest rates, work for about 
US$ 1 per day, not to benefit from subsidies, and become effective stewards of 
natural resources amidst poor transport infrastructure and weak support institu- 
tions. This is in sharp contrast to the situation faced by farmers in the North, who 
borrow money at low interest rates, have higher opportunity cost of labor, con- 
tinue to benefit from deeply institutionalized subsidies (e.g., for fertilizer, cereals, 
milk) and many cost-sharing schemes (when positive environmental externalities 
exist), and rely on a superb transport and information infrastructure. An enabling 
policy environment to the farming sector in developed countries is no justifica- 
tion for transferring the same elsewhere; but African farmers could really use a 
break in replenishing one of their most basic of natural resources — the soil. 


The authors are grateful to John Lynam, Paul Vlek, Ray Weil, Peter Cooper, 
Steve Franzel, and Cheryl Palm for their helpful criticisms of earlier versions of 
this chapter. 


Abelson, P.H., and J.W. Rowe. 1987. A new agricultural frontier. Science (Washington, DC) 

Ae, N., J. Akihara, K. Okada, T. Yoshinara, and I !. Johansen. 1990 Phosphorus uptake by pigeon pea 
and its role in cropping systems in the Indian subcontinent. Science (Washington, DC) 

Anderson, J.R. (ed.). 1994. Agricultural technology: Polic\ issues for the international community. 
CAB Int., Wallingford, England. 

Ashby, J.A. 1986. Th ff ifdit'1 rent I p ot'1 nner parti ipation n tilt m na ement of on-farni 

trials. Agric. Admin. 22:1-19. 
>,!il I i . letbodolos for the participation i mail farmei in th< lesitin o1 n inn trials. 

Agric. Admin. Ext. 25:235-252. 
in I.B. 19' I il and economical req in m nts f istainabie land use in si b ib i in 

Africa. Forum Dev. Stud. 2:211-219. 

Badiane, O., and C.L. Delgado. 1995. A 2020 vision for food, agriculture and the environment in sub- 
Saharan Africa. Food, Agric. and the Environ. Dis. Pap. 4. Int. Food Policy Res. Inst., 
Washington, DC. 

Balasubramanian, V.. and N.K.A. Blaise. 1993. Short season fallow management for sustainable pro- 
duction in Africa, p. 279-293. In J. Raglancl and R. Lai (ed.) Technologies for sustainable agri- 
culture in the tropics. ASA Spec. Publ. 56. ASA, CSSA, and SSSA, Madison, WI. 
'Barrios, E., R.J. Buresh, and J.I. Sprent. 1996. Nitrogen mineralization in density fractions of soil 
ii i i i n i i i n I I n i ii in I I i I i i l-l I 

Bationo, A., S.H. Chien, J. Henao, C.B. Christianson, and A.U. Mokwunye. 1990. Agronomic eval- 
ii ii in >1 i ] in ill i luil ited pbosphal ks in Ii- in us to Xi i il Sci. Soc. Am. J. 


Bationo. A., and A.U. Mokwunye. 1991. Alleviating soil fertilitx constraints to increased crop pro- 
duction in West Africa: The experience in the Sahel. p. 195-215. In A.U. Mokwunye (ed.) 
Alleviating soil fertility constraints to increased crop production in West Africa. Kluwer 
Academic Publ.. Dordecht. the Netherlands. 

F i lughogho. and A.U. Mokwunye. 1 i mi luatioi f ph pi i i i I 

in i i I 1 i i i I * \ < vl in nd P.L.G. Vlel 1.) Mai n in i 

i ii n ii I "I L I ii i n 1 l i in 1 1 laran Afri lartinu ijhoff. Dordrecht the 



Bationo, A., E. Rhodes, E.M.A. Smaling, and C. Visker. 1996. Technologies for restoring soil fertili- 
ty, p. 61-82. In A.U. Mokwunve el al. < eel. i Restoring and maintaining the productivity ot 
Y> e ;t African soils: Key to sustainable development. Misc. Fert. Stud. 14. Int. Fert. Dev. Ctr„ 
Africa, Lome. Togo. 

Bekunda, M.A., A. Bationo, and H Ssali 191 7 li] i rtilitj i i in ._ mi nl in Africa: A review of 
I i rcli trial ( '// R.J. Bu li et al. (eel.) Repleni nu i n i tili i in 

Africa. SSSA Spec. Publ. 51, SSSA, Madison, WI (this publication). 

Bikienga, I.M. 1997. Strategy tor large-scale use of Burkina phosphate. In H. Breman (ed.) 
International workshop on file development ot national strategies tor soil fertilitv recapital- 

tion i i I n i n ica inch tin ih i i i i h \ h i ' nd othei nendmeni 

II 1 " Mi a and World Bank. Int. Fert. Dev. Or.. Africa. Lome. Togo (in press). 

Black, A.S., and S.A. Waring 1976. Nitrate leaching and adsorption in a Krasnozem from Redland 
Bay, Queensland. Aust. J. Soil Res. 14:171-188. 

Borlaug, N.E., and C.R. Dowswell. 1994. Feeding a human population that increasingly crowds a 
111 te lecture, i ipplemei Ti i I il Id C'oi I 

Science, Acapulco, Mexico. 10-16 July 1994. ISSS, Wageningen, the Netherlands. 

Breman, H. 1990. No sustainability without external inputs, p. 124-134. In sub-Saharan Africa 
n i lju M i i li in i ii i tricht. Pro) roup Afri Directorat neral foi 

Int. Coop., Ministn if Foreign \u ii tin 1 1 gu th -etherlands. 

Eirouwer. J., and ''Mi li I 1 " i oil aspect i ii nl ling in a n mi ppli tion peri 

ment in Niger, p. 211-226. In J.M. Powell et al. (ed.) Livestock and sustainable nutrient 
1 ii hi mixed 1 nun \ i m f sul laran Frica I. II. Tech. Pa] hit Li tocl 
Ctr.-Africa, Addis Ababa, Ethiopia. 

Bumb, B.L., and C.A. Baanante. 1996. The role of fertilizer in sustaininsj food security and protect- 
ing the environment to 2020. Food, Agric. and the Environ. Dis. Pap. 17. Int. Food Policy Res. 
Inst., Washington, DC. 

Buol, S.W., PA. Sanchez, S.B. Weed, and J.M. Kimble. 1990. Predicted impact of climatic warming 
on soil piopeitie ind use. p 1 _ )ni| i i li id ti n I limat 

change on global agriculture. ASA Spec. Publ. 53. ASA, CSSA, and SSSA, Madison, WI. 

Buol. S.W. and M.L. Stokes. 1997. Soil profile alteration under long-term hi till input agriculture, p. 
17-109. In R.J. Bureshetal I.) Repleni hm I fertility in Afri | ul 1 

SSSA, Madison, WI (this publication). 

Buresh, R.J., PC. Smithson, and D.T. Heliums. 1997. Building soil phosphorus capital in Africa, p. 
I I i I i I. Buresh 1 l I i i I In i rica A Spec. Publ. 51. 

SSSA, Madison, WI (this publication). 

Buresh, R.J., and G. Tian. 1997. Soil impn vemenl I l sub haran Africa. Agrofor. Syst. (in 

Cahn M.D., D.R. Bouldin, and M.S. Cravo. 1492. Nitrate sorption in the profile of an acid soil. Plant 
Soil 143:179-183. 

i i n I., and 'ii I ' h ii M i hi in nu i i New lesearch atten- 

ii n i ,i in i i ' hn qu In R.J. Buresh and P.J. < per i In n ind prai 

i t i il i ii i talk I wer Acad mi 'ubl i rdi lit. th therland in 

Christianson, C.B nil" Mil i in fertilit traints to inci i 

1 II I 111 t V 11 L I 1 I II I i 1! Mill 1 I I I I I ' 

A. I". Mokwunye (ed.) Alleviating soil fertility constraints to increased crop production in 
est Africa In i iden ubl.. Dord it. th i rlan 

Cleaver, K. 1993 A t a . ti levelop igri ulturt in u ' iharan rt'rii mi focu foi u, <'■ nid 
Bank. World Bank Tech. Pap. no. 203. World Bank, Washington, DC. 

Cleaver. K.M.. ami C.A. Schreiber. 1994. Reversing the spiral, the population, agriculture and envi- 
ronment nexus in sub-Saharan Africa. World Bank, Washington, DC. 

Conway. G.. and L. Barbier. 1990. After the green revolution: Sustainable agriculture for develop- 
ment. Earthscan Publ.. London. 

Crosson. P.. and J.R. Anderson. 199:3. Achieving a sustainable agricultural svstem in sub-Saharan 
Africa. Building blocks for Africa 2025. Pap. 2. World Bank, Washington, DC. 

Crowley, EX. 1996. Off-farm income and farming in western Kenya. Rep. to USAID/Kenya, Nairobi, 
Kenya. TSBF, Nairobi, Kenya. 

Davis. T.J.. and I mini i i I 1987 i inn litv issu in icultural d \elopment. Proc. of 

the Agriculture Sec tor Symp., 7th, Washington, DC. 8-9 Jan. 1987. World Bank, Washington, 

De Jager, A., and E.M.A. Smaling. 1996. Turning the tide, p 83-94. In A. I ! . Mokwunye et al. (ed.) 
Restoring and maintaining the productivity of West African soils: Key to sustainable devel- 
opment. Misc. Fert. Stud. 14. Int. Fert. Dev. Ctr.-Africa. Lome. Togo 


Delgado, C, J. Hopkins, and V. Kelly. 1994. Agricultural growth linkages in sub-Saharan Africa: A 
nthesis. ] '1 rl [ -> i ilti I ' i hi mi ii 

Africa, Washington, DC. 26 May 1994. Int. Food Policy Res. Inst., Washington, DC. 
D ml I 1 X Iiiii h i n 1 in i 1 i i i hi i i tilii r In lo I lopment and imple- 

mentation in ill i Aln th i 1 i n i il I ii In i in in nl mm 

tive in Burkina Faso. Int. Fert. Dev. Ctr.-Africa, Lome, Togo. 
i i I I > i ii i ii i n I i i ' i i in Mi hungry world, p. 

' I i i c ii n 1 1 ' i i in n Fraud C'A lar. x inn 

Agric. Univ., Wageningen. the Netherlands. 
DeWit, C.T. 1992. Resource efficiency in agriculture. Agric. Syst. 40:121-151. 
Dommen, A.J. 19! Ini \ tion n African agriculti Vestview ! ! ikler. CO 

Donovan, W.G 199 I it I i npui ml m i i in ii th i ' n ition of agriculture. 

p. 178-194 In $ i Hi c i lit n i ii im i i n i u i rch invest nts in rica 

Sasakawa Africa Assoc, Mexico City. 
Eicher, C.K. 1982. Facing up to Africa's food crisis. Foreign Affairs 61 (1):154— 174. 
FAO. 1971. Improving soil fertility in Africa. FAO Soils Bull. 14. Food and Agric. Organ, of the 

United Nations, Rome. 
FAO. 1988. Revised legend of the FAO-UNESCO soil map of the world. Food and Agric. Organ, of 

the United Nations, Rome. 
FAO. 1995. FAO fertilizer yearbook 1994. Vol. 44. Food and Agric. Organ, of the United Nations, 

FAO. 1996. Soil fertility initiative. Food and Agric. Organ, of the United Nations, Rome. 

Fisher M.J., I.M. Rao, MA. Ayarza, C.E. Lascano, J.I. Sanz, R.J. Thomas, and R.R. Vera. 1994. 

Carbon i I ntroduced dee[ I n the St ml neri i i n i iti 

(London) 371:236-238. 
Franzel, S. 1998 Socioeconomi fa to al acting tin idoptioi ofpol ntial improved tree fallows, fa 
I F tn i ii i 1 I ' t i I i 1 in ml i n ii i put in im; \ i lit \ 

Khmer Academic Publ.. Dordrecht, the Netherlands (in press). 
Fresco L.O., and S.B. Krooneiibers. 1992. Time and spatial scales in ecological sustainabilitv. Land 

Use Policy. 9:155-167. 
Frissel M.J. (ed.). 1978. Cycling of mineral nutrients in agricultural ecosystems. Elsevier, 

Amsterdam, the Netherlands, 
i ii i 1 VI. B rd. M..I. Hedi nd > Vleth II I I tion t nn illing the cycling of 

P in soils, p. 107-137. /» H. 1 It II 

cycles and management. Scope 54 John A'il oi New York. 

Gachengo, C.N I > II III 'I i i niitenal to phos- 

phorus fixing soils. M.S. thesis. Moi Univ., Eldoret, Kenya. 
Galiba, M., P. Vissoh, G. Dagbenobakin. and F. Fagbohoun. 1997. Reflexions et apprehensions 

paysannes lies a la vulgarisation du pois mascate (Muciiiiu ulili.s I an Benin. Sasakawa Africa 

Association, Mexico City. 
Garrity D.P. 1996. Tree-soil-crop interactions on slopes, p. 299-318. In C.K. Ong and PA. Huxley 

(ed.) Tree-crop interactions. A ph\ si. i i ippi 1 ' ' Im >' llin ird England. 
Garrity, D.P, and A.R. Mercado, Jr. 1994. Nitrogen-fixation capacity in the component species of 

contour hedgerows: How important? Agrofor. Syst. 27:241-258. 
Giller, K.E., and G. Cadisch. 1995. Future benefits from biological nitrogen fixation: An ecological 

approach to agriculture. Plant Soil 174:255-277. 
filler, K.E., G. Cadisch, C. Ehaliotis, E. Adams, W.D. Sakala, and PL. Mafongoya. 1997. Building 

soil nitrogen capital in Africa, p. 151-192. In R.J. Buresh et al. (ed.) Replenishing soil fertil- 
ity in Africa. SSSA Spec. Publ. 51. SSSA, Madison, WI (this publication). 
Giller, K.E., and K.J. Wilson. 1991. Nitrogen fixation in tropical cropping systems. CAB Int., 

Wallingford. England. 
Gladwin, C.H., K.L. Buhr, A. Goldman, C. Hiebsch, P.E. Hildebrand, G. Kidder, M. Langham, P. 
kedi-Kizza md D William 199 Gender an >il fertility in Afri i p 219-236 In R.J. 

Buresh etal. (eel. ! Replenishing soil fertility in Africa. SSSA Spec. Publ. 51. SSSA, Madison, 

WI (this publication). 
Goedert, W.G. 1985. Solos dos cerrados Tecnologi trategi de manejo Nobel Sao Paulo, 

Goedert, W.G. 1987. Management ol acid tropical soils in the savannas .if South America, p. 

109-127. In Management ol acid tropical soils for sustainable agriculture. IBSRAM Inaugural 

Woikshop. Int Board fo i il R ml Vlanag i mgl ik Thailand. 
Goedert, W.G., and E. Lobato. 1980. Eficiencia agronomica tie fosfatos em solos de Cerrado. Pesqui. 

Agropecu. Bras. 15:311-318. 


Hammond, L.L., S.H. Chien, and A.U. Mokwunye. 1986. Agronomic value of unacidulated and par- 
tially acidulated phosphate rocks indigenous to tiie tropics. Adv. Agron. 40:89-140. 

M nd Ml i Harri n nidi I i n mitli 199 Plios] on lami m slash-and-burn and 

II i i pin r in i tn I m i I t ] ] i i i i ii li hi n i 

1 ' 1 i ii urn i I i i ] ii i i i n in pe 54. John Wil n 

1 i hi I i I mi I lama, ami B.H. Janssen. 199 il n I w ater dynamics in 

ii bania fallo ■ edfallow and maiz Soil Set. Soc. Am. J. 60:568-574. 

Hedley, M.J., J.J. Mortvedt, N.S. Bolan, and J.K. Syers. 1995. Phosphorus fertility management in 
agroecosystems. p. 59-92. In H. Tiessen (ed.) F'hosph ru in th. ilobal environment: 
[Yanst'ei It nd management lohn i on I V'orl 

H n ii nn i. 19 | i lection for all pin n rn I n tent nianagement. 

ii i u iii ill i n I ii 1 19 [). diss. Oxford Univ., 

Oxford. England. 

Heisey, P.W., and \\ M isii. 1996. Fertilizer u I mai roducti urn il ihaian Africa. Econ. 

Work. Pap. 9 6 - 1 . CIMMYT, Mexico City. 

Hoekstra. D. 1988. Summar) of the zonal agroforestrv potentials and research across land use sys- 
tems in the highlands of eastern and central Africa. AFRENA Rep. 5. Int. Ctr. for Res. in 
Agrofor., Nairobi, Kenya. 

Hoekstra, D., and J.D. Corbett. 1995. Sustainable agricultural growth for the highlands of East and 
Central Africa: Prospects to 2020. Int. Food Policy Res. Inst., Washington, DC. 

Holden, S.T., and N. Shanmugarathan. 1994. Structural adjustment, production subsidies and sus- 
tainable land use. Dis. Pap. D-l 1/1994. Dep. of Econ. and Social Sci., Agric. Univ. of Norway, 

Homer-Dixon T.F., J.H. Boutwell, and G.W Rathfens. 1993. Environmental chanse and violent con- 
flict. Sci. Am. 268 (2): 16-23. 
IAEA. 1975. Root activity patterns of some tree crops. Tech. Rep. Ser. no. 170. Int. Atomic Energy 

Agency, Vienna, Austria. 
ICRAF. 1995. Annual report 1994. Int. Ctr. for Res. in Agrofor., Nairobi, Kenya. 
ICRAF. 1996. Annual report 1995. Int. Ctr. for Res. in Agrofor., Nairobi, Kenya. 
ICRAF. 1997. Annual report 1996. Int. Ctr. for Res. in Agrofor., Nairobi, Kenya. 
IFPRI. 1995. A 2020 vision for food, agriculture and the environment: The vision, challenge and rec- 

omiiiended action. Int. Food Policy Res. Inst., Washington, DC. 
IFPRI. 1996. Feedi I j 1 I .1 

Policy Res. Inst., Washington, DC. 
ETA. 1995. Annual report 1994. Int. Inst, of Tropical Agric, Ibadan, Nigeria. 
Ikerra, T.W.D., P.N.S. Mnekeni, and B.R. Singh. 1994. Effects of added compost and farmyard 

manure on P release from Minjhiiiu phosphate rock and its uptake bv maize. Norw. J. Asnlc. 

Sci. 8:13-23. 
Iyamuremye, F., and R.P. Dick. 1996. Organic amendments and phosphorus sorption b\ soils. Ad\ 

Agron. 56:139-185. 
Izac. A-M.N. 1997. Ecological economics oi investing in natural resource capital in Africa, p 

237-251. In R.i Buresh et ah (ec lishing I fertility in Africa. SSSA Spec. Publ. 51. 

SSSA, Madison, WI (this publication). 
Kabaara. A.M. 1064. Nitrate studies in a fertilizer experiment in .Sukumaland Tanganyika. East Afr. 

Agric. For. J. 30:142-148. 
1 kiyai. J. I' 96 Dynami of soil oi ni irl n niti n m I m bial bioma i long-l nn 
ii en n i I I m i i i m i i Ii i \1 thesis. D l il 

Univ. of Nairobi, Nairobi, Kenya. 

! 1 I I I lations. Vol. 1-22. Fertil 

Use Recommendation Project. Kenya Agric. Res. Inst., Nairobi, Kenya. 
Kiepe, P., and M.R. Rao. 1994 Manas; m ni >i irofoi ti for th • on i u m and utilization of 

land and water resources. Outlook Agric. 23 (1):17— 25. 
Kinjo, T, andP.F. Pratt. 1971. Nitrate adsorption. Soil Sci. Soc. Am. Proc. 35:722-732. 
Kirk, G.J.D., and P.H. Nye. 1986. A impk mi ' I i red n th Kites of dissolution of sparingly 

soluble calcium phosphates in soil. J. Soil Sci. 37:529-554. 
Kwesiga, F.R., and M. C'hisumpa. 1992. Ethnobotanical surve\ in eastern province Zambia. AFRE- 
NA Rep. 49. Int. Ctr. for Res. in Agrofor., Nairobi, Kenya. 
Kwesiga F.R., and R. Coe. 1994. The effect of short rotation Sesbania sesban planted fallows on 

maize yields. For. Ecol. Manage. 64:199-208. 


Kwesiga, F.R., S. Fran I. F. Pi D. Phiri. an I ini n i I ia sesban improved 

i I i eastern mbi fheir inceptioi dt I | in nt n I ntl u ism 1 I 

Bui ncl ! I ' ' | i 'II ieix no practi f short-t n im id d t'allo\ lu\ i 

Academic Publ.. Dordrecht, the Netherlands (in press). 
Lajtha, K, and A.F. Harrison. 1995 Stratc i n Ik pli in cqti ition and conservation by plant 
i i i inn i mil 139-1 ' 1 1 I i ii i ' ii i u i i 1 i I in i 

ii i 1 1 i 1 i I ii I ii i n i i i -I hn ii Sons New York. 

II ill I i n I n i ii i I I i I i 1 1 i ettect An overview. 

p. 1-7. In R. Lai et al i i ind global chan Lewi I nbl Boca Raton, FL. 

Larson. B A., and G.B. Frisvokl. 1996. Fertilizers to support agricultural development in sub-Saharan 

Africa: What is needed and why. Food Policy 21:509 — 525. 
Lele. U. ! ( J8!. Rural Africa: Modernization. equit\ and Ions; term development. Science (Washington. 

DC) 211:557-583. 
Lompo, F. 1993. Contribution a la valorisation des phosphates natureles tin Burkina Faso: Etudes des 

effets de l'intei tion } losphati natui I -nun i uganiqin ; Tli Docteui Ingenieur. 

Faculte des Sciences el Techniques de L'Universite Xationale de Cole d' Ivoire. Abidjan. 
I j % I I i i ' i i ii | i I ill ' de Potassa e 

Fosfato, Piracicaba, S.P., Brazil. 
Lopes, A.S. 1996. Sods under Cerrado: A success story in soil management. In IPA-PPI Regional 

Conf. for Latin America and the Carribean, Mexico City. 25-28 June 1996, IPA-PPI, Mexico 

Lopes, A.S. , and L.R.G. Guilli rn 1994 ,i () l rradi dm jo d i 1 1 1 d ide para producao 

agropecuaria. Boletim Tecnico No 5, 2a edic3o. ANDA, Sao Paulo, Brazil. 
Manyong, V.M., A.V. Houndek. i 4 Grogan vl l < rsteeg and F. van der Pol. 1997. Determinants 
! ii i i i n i i i I I una in Benin F i 

1-86 — ' / Ii i i i I in agricultural and biol ii nn 

engineering. China Agric. Univ. Press, Beijing. 
McClellan, G.H., and A. Notholt. 1986. Phosphate deposits of tropical sub-Saharan Africa, p. 

I 4. / I kwui ill VI i in ii n n n i i i ii In 

fertilizers in sub-Saharan Africa. Martinus Nijlioff. Dordrecht, the Netherlands. 
Mclntire, J., andJ.M. Powell. 1995. African semi-arid tropical agriculture cannot grow without exter- 
1 i i In J.M 11 i I Livestock and sustainable nutrient in 

I I fuel Vol. II. Tech. Pap. Int. Livestock Ctr. for 

Africa. Addis .Ababa, Ethiopia. 
Mekonnen, K., R.J. Buresh, and B. Jama. 1997. Root and inorganic distributions of sesbania fallows, 

natural fallow and maize. Plant Soil 188:319-327. 
il I i ii I laclntyre. I Phosphoi ncentratioi i 1 I limitation in tropical 

.hi rs and lal I InH.Ti u and 1 i i I Pi I i 1 u i i 

i I | i i Regional Workshop. 4th. Nairobi. Kenya. I i 

1991. Saskatchewan Inst, of Pedology. I'niv. of Saskatchewan. Saskatoon, Canada. 
Mills, WR. 1953. Nitrate accumulation in Ugandan soils. East Afr. Agric. For. J. 19:35-36. 
Mokwunye, A.U. (ed.). 1991 I I ili nstraints to inci u op production in 

West Africa. Kluwer Academic Publ.. Dordrecht, the Netherlands. 
Mokwunye, A.U., A. de Jager, and E.M.A. Smaling (ed.l. 1996. Restoring and maintaining the pro- 
i y of Wesl in n oil . to ii i in i velopmeni Vlisi F t tud 14 Int. Feit. 

Dev. Ctr.-Africa, Lome, Togo. 
Mokwunye, A.U., and L.L. Hammond. 1992. Myths and science of fertilizer use in the tropics, p. 

121-134. In R. Lai and PA. Sanchez (ed.) Myths and science of soils of the tropics. SSSA 

Spec. Publ. 29. SSSA, Madison, WI. 
Mokwunye A LT in L i I I nt of nil nd pi phorus fertilizers in 

sub-Saharan Africa. Martinus Nijlioff. Dordrecht, die Netherlands. 
Mugwira, L.M., and L.M. Mukurumbira. 1986. Ntui ni ippl in i ditteient groups of 

manure from communal areas and commercial feedlots. Zimbabwe Agric. J. 83:25-29. 
Murwira. K H.. M.J. Swift, and P.G.H. Frost. 1995. Manure as a key resource in sustainable agricul- 
ture, p. t — 48. /// J.M. Powell ei al. (ed.) Livestock and sustainable nutrient cycling in mixed 

farming systems of sub-Saharan Africa. Vol. II. Tech. Pap. Int. Livestock Ctr. for Africa. Addis 

Ababa, Ethiopia. 
Nagarajah, S.. and L marasiri. I I rganic mal I tilizei i lowland rice in Sri 

Lanka, p. 97-104. In Soil o nil matt i tudi I nt I mil Enei *. nc\ Vienna, Austria. 
Nagarajah, S., and B.M. Nizar. 1982. Wild sunflower as a green manure for rice in the mid country 

wet zone. Trop. Agric. 138:34-40. 


Narayan. IX, and D. Nyaimvaya. ! ! 
pared for the World Bank ai 
AMREF, Nairobi, Kenya. 

Ngugi, D.N. 1988. Agroforesti research j | i th nai li\ t k i m in the unimodal 
upland platea i 1 istern Pi nee (Chipal ml Katel listricts) ni'bi N I > \ Rep 10. 
Int. Ctr. for Res. in Agrofor.. Nairobi. Kenya. 

Niang, A., B. Amadalo, and S. Gathumbi. 1996. Green manure from the roadside, p. 9-10. In Miti Ni 
i n i I Vol \ i ill mu Kenya. 

Nye, P.H., and D.J. Greenland. 1960. The soil under shifting cultivation. Commonwealth Bur. of Soils 
Tech. Commun. 51. Commonwealth Agric. Bur.. Farnhani Royal, Bucks, England. 

Nziguheba, G., C.A. Palm, RJ. Buresh, and P.C. Smithson. 1998. Soil phosphorus fractions and sorp- 
tion as affected by organic and inorganic sources. Plant Soil (in press). 

Okalebo, J.R., and R L K oomei I' 1 Phosphoru re ourct \p]i iton pi j i toi western Kenya 
(PREP). Trap. Soil Biol, and Fert. Prog., Nairobi, Kenya. 

Osei-Bonsu, P., and D. Buckles. 1993 Com Hi Is and improvii il rtility through the use 

i i i i i n ii 1 ; in Benin and Chan In West Africa 

nun i I I i! 14. Int. Inst, of Trap. Agric. Ibadan. Nigeria. 

Oswald, A., H. Frost, J.K. Ransom, J. Kroschel, K.D. Shepherd, and J. Sauerborn. 1996. Studies on 
n i i i ' f in I il n i i i limbs to red u nfestatioi n Kei 

In M.T. Moreno et al. (ed.) Proc. of the Int. Parasitic Weed Symp., 6th, Cordoba, Spain. 16-17 
Apr. 1996. Junta de Aiulalucia. Cordoba. 
I ii 1 ' nti i i ii n i ( i i i ii i i in i in q hi i nt t im i 1 p J ' | ' "" '■ 

Agrofor. Syst. 30:105-124. 

Palm, C.A., J. Mukalama, J. Agunda, P. Nekesa. S. Ajanga. and J. Odhiambo. 1997a. Farm hedge sur- 
vey: composition, iiianageinent. use and potential for soil fertility management. Trop. Soil 
Biol, and Fert. Prog., Nairobi, Kenya. 

Palm. C.A.. R.J.I \\ t ml mdw 19 b ' mbin in toi u md inorganic nutri- 

ent sotu for soil fertility ma intenan. ami replenishment, p. 19 In R..I. Bui i i il 

(ed.) Replenishing soil fertilitj in Africa. SSSA Spec. Publ. 51. SSSA, Madison, WI (this pub- 

Penning de Vries, F.W., and MA. Djiteye (ed.). 1982. La productivite des paturages saheliens. Un 
elude de sols des \eget in id! ploitation d ette i ui i iturells. PUDOC, 
1 mu i Ui Hi il n 1 

Pieri, C.J.M.G. 1989. Fertilitj of soils. A future for farming m the West African savannah. Springer- 
Verlag, Berlin. 

Probert, M.E., J.R. Okalebo, and R.K. Jones. 1995. The use of manure on smallholders' farms in 
semi-arid eastern Kenya. Exp. Agric. 31:371-381. 

Quiflones, M.A., N.E. Borlaug, and C.R. Dowswell. 1997. A fertilizer-based green revolution for 
Africa, p SI Mm Ii 1 i plenishin il i i Atncr SSSA Spec. 

Publ. 51. SSSA, Madison, WI (this publication). 

Quiflones, M.A., and Takele Gebre. 1996. An overview of the Sasakawa Global 2000 project in 
Ethiopia p 1 1 I \ ' rem h ' n n reater impact from i rch inve tments in 

i i ii 1 i i i n 

Qureshi. J.N. 199 I Tin- cumulative effects of X-P fertilizers, manure ami crop residues on maize and 
bean yields and some soil hemical properti i i Ml ' I'ecent advances in 

Mi h i nun 1 i ' I i un I i ni i i i 1 ii i 

Sept. 1990. Kenya Agric. Res. Inst., Nairobi. Kenya. 

Reinjtjes, C, B. Haverkrot, and A. Waters-Bayer. 1992. Farming fin the future. An introduction to 
low external input and sustainable agriculture Macmillan Press, London. 

Ribeiro. J.L. 1979. Taxas de retorno an in\ estiinento agricola na regiao dos C'errados no Brasil. p. 
I F i i i i ii i i n 1 i ! i i I i u inejo V Simposio 

sobre o Cerrado. EDITERRA, Brasilia, Brazil. 

Runge-Metzgei \ I losing tl mi Obsl i i ill nt P n n i t toi improved glob- 

al food security, p. 27-42. In H. Tiessen (ed.) Phosphorus in the global environment: 
i ii i I I i i in i i p I ii NI ) 1 

liez. P.A. I t i ill ii il i i i I \ hit i i i 

n i I 1994. Tropical il fertility i rch. towards th ond paradigm. [ . Fran 

I i ' orld Congi i il i in -, ml . I n. In i lul 1994. I ' nigei 

the Netherlands. 

Sanchez, P.A. 1995. Science in agroforestry. Agrofor. Syst. 30:5-55. 


Sanchez, P.A., and JR. Benites. 14X7. Low input cropping for acid soils of the humid tropics. Science 

(Washington, DC) 238:1521-1527. 
Sanchez, P.A., and S.W. Buol. 1975. Soils of the tropics and the world food crisis. Science 

(Washington, DC) 188:598-603. 
Sanchez, P.A., R.J. Buresh, and R.R.B. Leakey. 1997 Trees, soils, and food security. Philos. Trans. R. 

Soc. London B 352:949-961. 
Sanchez P.A., W. Couto, and S.W. Buol. 1982. The fertilitj capability soil classification system: 

Interpretation, applicability ami modification. Geoderma 27:283-309. 
Sanchez, P.A., A-M.N. Izac, I. Valencia, and C. Pieri. 1996. Soil fertility replenishment in Africa: A 
ii i i i I hi i iter impact fn in i i li in i 

nents in Afri I i lexico C'ii 

i» li [ nd R.R.B. Leal I 1 I mil i ransform iti n in in i I In let nniii i lit i >i 

balancing food security with natural resource utilization. Eur. J. Agron. 7:l- ( ). 
Sanchez, P.A., and C.A. Palm. 1996. Nutrient cyclhuj and aiiroforestry in Africa. Unasylva 185 

Sanchez. P.A.. and J.G. Salinas. 1981. Low-input technology for managing Oxisols and Ultisols in 

tropical America. Adv. Agron. 34:280-406. 
Sanchez. P.A.. and G. I'ehara 1980 Management considerations Tor acid soils with high phosphorus 

fixation capacity, p. 471-514. hi F.E. Khasawneh et al. (ed.) The role of phosphorus in agri- 
culture. ASA, CSSA, and SSSA, Madison, WI. 
Sanders. J.H.. B.I. Shapiro, and S. Ramaswamy. 1996. Tile economics of agricultural technology in 

semi-arid sub-Saharan Africa. Johns Hopkins I niv. Press. Baltimore. V1D 
Serageldin, I. 1995. Sustainability and the wealth of nations. World Bank, Washington, DC. 
Sharpley AN M I II II l ibb n Hill ht-l ki ka Hi u md L Ryszkowsky. 

1995. Phosphorus transfers from terrestrial to aquatic ecosystems, p. 171-199. In H. Tiessen 

(ed.) Phosphorus in the global environment transfei cvcl md manaa ineni o 4 loin 

Wiley & Sons, New York. 
Shepherd, K.D., and M.J. Soule. 199 nil n mi n i i al impacts ofagro- 

I (I 1 li I I 

lation model. Agin Eci \ Environ, (in press). 
Smaling, E.M.A. 1993. An agroecological framework for integrating nutrient management, with spe- 
cial reference to Kenya. Ph thesi ric. LTi nin n i \etheilands 
Smaling, E.M.A., S.M. Nandwa, and B.H. Janssen. 1997. Soil fertility in Africa is at stake, p. 47-61. 

/[If' i et al. (ed leni Inn ul 1 n b n frii Publ I 

Madison, WI (this publication). 
i i I I i I i I i ' | i I i i i I ii I in Cerrado soils as 

affected by phosphorus sorption capacity. Soil Sci. Soc. Am. J. 46:339-345. 
il 1 1 'ii 1 i i i in ii il i i ii j n i I i i 

19. Pocahontas Press. Blacksburg, VA. 
Ssali, H., P. Ahn, and A.U. Mokwunye. 1986. Fertility of soils in tropical Africa: A historical per- 
pectix > i i I ii i l I I lek (ed.) Management niti i i I 

phosphorus fertilizers in sub-Saharan Africa. Martinus Nijhoff. Dordrecht, she Netherlands. 
Stoorvogel, J.J., and E.M.A. Smaling. 1990. Assessment ot soil nutrient depletion in sub-Saharan 

Africa: 1983-2000. Rep. no. 28. Vol. 1-4. Winand Staring Ctr., Wageningen, the Netherlands. 
Swift, M.J., P.D. Seward, P.G.H. Frost, J.N. Qureshi, and F.N. Muchena. 1994. Long-term experi- 
ments in Africa Iv li ping d lab i for sust inable I m i u undt i do , ! hangt ) 

229-251. /;; RA. Leigh and A.E. Johnson (ed.) Long-term experiments in agricultural and 
i I ii i I in i I 1 1 ii nun 'i I England. 
Swinkels, R.A., S. Franzel, K.D. Shepherd, E. Ohlsson, and J.K. Ndufa. 1997. The economics of short 

rotation improved fallows: Evidence from areas of high population density in western Kenya. 

Agric.Syst. 55:99-121. 
Szott, L.T., C.A. Palm, and R.J. Buresh. 1998. Ecosystem fertility and fallow function. In R.J. Buresh 

and P.J. Cooper (ed.) The science and practice of short-term improved fallows Kluwer 

Academic Publ., Dordrecht, the Netherlands (in press). 
Szott, L.T., C.A. Palm, and P.A. Sanchez. 1991. Agroforestrj in acid soils of the humid tropics. Adv. 

Agron. 45:275-301. 
Tinkei P.E 19 i oil chemistry of phosphoru md mycorrhizal effect on plant growth, p. 353-371. 

In F.E. Sanders et al led.i Endomycorrhizas. Academic Press. London. 
Tomich, T.P., P. Kilby, and B.F. Johnston. 1995. Transforming agrarian economies: Opportunities 

siezed, opportunities missed. Cornell Univ. Press. Ithaca, NY. 


Van der Pol. F. 1992. Soil mining: An unseen contribution to farm income in Southern Mali. Bull. 325. 

Royal Trop. Inst.. Amsterdam, the Netherlands. 
Van Kauwenbergh, S.J., L.B. Williams, and G.H. McClellan (ed.) 1991. The fertilizer mineral 

n 1L1 nd i in i in Spe i ,1 issue. Fert. Res. 30:123-296. 

Wang'ati, F., and K. Kebaara (ed.). 1993. Integrated natural resource management research for the 

highlands of East and Central Africa Int. Or. for Res. in Agrofor.. Nairobi. Kenya. 
White, F. 1983. The vegetation of Africa. United Nations Educ, Sci. and Cultural Organ., Paris. 
Woomer, P.L., J. Kotto-Same, MA. Bekunda, and J.R. Okalebo. 1997. The biological management of 

tropical soil fertility: Some research and development priorities for Africa. Proc. Long-Term 

Soil I i n [in i I I i te Univ.. Columl 'Him 

Woouicr. P.L.. and FN. Muchena. 1983. Recognizing and overcoming soil constraints to crop pro- 
duction in tropical Africa. Afr. Crop Sci. .1. 4:5(13-518. 
World Bank. 19S9 i harai u rom crisis I ustainabl rowth A long term perspective 

study. World Bank, Washington, DC. 
World Bank. 1990. World Bank development report 1990. World Bank, Washington, DC. 
v' >rld Banf !-' Toward environmental!; ustainable d elopn nt in ub ialiaran Africa \ World 

Bank agenda. Rep. no. 15111-AFR. World Bank, Washington, DC. 
World Bank. 1996a. African development indicators 1996. World Bank, Washington, DC. 
World Bank. 1996b. Natural resource degradation in sub-Saliaran Africa: Restoration of soil fertility. 

Africa Region. World Bank, Washington, DC. 
Wortmann, C.S., M. Isabirye, and S. Musa. 1994. Crotolaria as a green manure in Uganda. Afr. Crop 

Sci. J. 2:55-61. 
Yates, R.A., and A. Kiss. 1992. Using and sustaining Africa's soils. Agric. and Rural Development 

Ser. 6. World Bank, Washington, DC. 
Yost, R.S., E.J. Kamprath, E. Lobato, and G. Naderman 1979. Phosphorous response on anOxisol as 

influenced by rates and placement. Soil Sci. Soc. Am. J. 43:338-343. 
Yost, R.S., G.C. Naderman, E.J. Kamprath. and E. Lobato. 1982. Availability of rock phosphate as 

measured by an acid-tolerant pasture grass and extractable phosphorous. Agron. J. 



Soil Fertility in Africa Is at Stake 

Eric M. A. Smaling 

Research Institute for Agrobiology and Soil Fertility 

Wageningen, the Netherlands 

Stephen M. Nandwa 

b • • i \ ', ; '/(i ' Reseat: In I In 
Nairobi, Kenya 

Bert H. Janssen 

Wageningen Agricultural University 

Wageningen, the Netherlands 


Soil fertility in Africa is under pressure as an increasing number of fanners attempt 
to make a living based on what the land can offer to growing plants. Studies in Africa from 
about 1989 have focused on difl i at p tial scales, i.e., subcontinental, subnational, and 
farm. This chapter reviews the results obtained at these three levels and compares method- 
ologies and implications. For N, annual depletion was recorded at all levels at rates of 22 
kg ha- 1 (sub-Saharan Africa), 112 kg ha- 1 (Kisii District, Kenya), and 71 kg ha" 1 (average 
for 26 farms in Kisii, Kakamega, and Embu Districts). If the soil nutrient balance is to 
become a suitable land quality indicator for wider use as a policy instrument, increased 
sophistication is required, including data on soil nutrient stocks and availability. The 
advantage of the nutrient balance approach over traditional rate-response research on fer- 
tilizers is that it includes all possible nutrient flows at the spatial scales discussed. A draw- 
back, however, is the lack of hard data on flows that are difficult to measure (leaching, 
gaseous losses, and erosion), and the fact that the balance comprises several inputs minus 
the sum of several outputs. Nonetheless, the message comes out clearly that improved soil 
nutrient management is crucial lor maintaining and improving soil productivity in Africa. 

Soil fertility is not a static feature. On the contrary, it changes constantly and its 
direction (accumulation or depletion) is determined by the interplay between 
physical, chemical, biological, and anthropog -nil proi esses. This dynamism also 
is reflected in terminology such as nutrient cycles, budgets, or balances, referring 
to inputs and outputs in natural ecosystems and managed agroecosystems, to 
which nutrients are added and from which nutrients are removed. As the world 


population keeps growing, balanced ecosystems are on the decrease and n 
ledgers all over the world have become increasingly imbalanced. Great nutrient 
surpluses and subsequent undesirable emissions to the environment now occur in 
many farming systems in temperate regions, and increasing soil-nutrient deple- 
tion and crop yield declines are reported in the tropics, particularly in rainfed sub- 
Saharan Africa (hereafter referred to as Africa; Pieri, 1989; Stoorvogel & 
Smaling, 1990; Van der Pol, 1992). In rural appraisals, an increasing number of 
African farmers indeed mention soil fertility decline as a major constraint to 

The yield-increasing effect of mineral fertilizers has long been the main 
nutrient management technology researched, amongst others, by the numerous 
though poorly documented rate-response trials of the FAO Fertilizer Programme. 
Presently, however, land-use planning approaches are aimed at integrated nutri- 
ent management (INM), perceived here as the best combination of available 
nutrient management technologies, i.e., those that suit local biophysical condi- 
tions and are economically attractive and socially relevant (Smaling et al., 1996). 
FAO has also adopted this philosophy and now runs an Integrated Plant Nutrition 
Programme in different parts of the tropics. Integrated nutrient management tech- 
nologies can be nutrient saving, such as in controlling erosion and recycling crop 
residues, manure, and other biomass, or nutrient adding, such as in applying min- 
eral fertilizers and importing feedstuffs for livestock. Some practices strive at 
both, such as improved fallowing and agroforestry. As each agroecological zone 
has its potentials and limitations, the number of relevant INM options is site spe- 
cific. In the eastern African highlands, for example, with reliable rainfall and 
deep, relatively fertile soils, more options are available to safeguard productivity 
than in semiarid West Africa, with less and erratic rainfall and sandy, often shal- 
low soils. Of late, the nutrient balance and INM have been adopted by the World 
Bank as key to the debate on sustainable agricultural systems in the tropics. As a 
consequence, work is under way to turn the nutrient balance into a land quality 
indicator (Pieri et al., 1995). 

Nutrient balances apply to different spatial scales. To visualize this, one 
should build an imaginary fence around the system of interest. For the farm sys- 
tem, for example, this fence surrounds the entire farm holding. The floor runsjust 
below the root zone of plant species that grow in the particular farming systems, 
whereas the roof stretches over the top of the tallest species. Now one can deter- 
mine whether a nutrient flow is really an input or an output, i.e., crossing this 
fence, or whether one deals with an internal flow inside the fence. Concentrates 
purchased to feed stalled cattle, for example, are nutrient inputs to the farm, but 
roughage such as napier grass (Pennisetum purpureum Schumach.) or silage 
maize (Zea mays L.) grown within the farm is no input at this level. It is, howev- 
er, an output for the plot where these plants were grown, and an input to the sta- 
ble, which are both compartments of the farm. Similarly, soil that leaves the farm 
through water erosion represents a nutrient output, but eroded soil from upper 
slopes may enter the same farm and become an input. Eroded soil reaching rivers 
may end up in the ocean and is also then an output at the country scale. Another 
percentage may, however, be deposited as sediment in flood plains in the lower 
parts of the river basin. This is the case in large parts of agricultural China, where 


soils in the plains remain productive by virtue of erosion in the mountains. At the 
level of the soil solution, the nutrient balance in fact represents plant-nutrient 
availability. Any applied P fertilizer is an input to the farm and the soil, but a large 
part may be strongly sorbed by sesquioxides or precipitated, and as such it is no 
immediate input to the soil solution. In other words, nutrient availability reflects 
a nutrient balance at soil-solution level, modified by the process groups mineral- 
ization-immobilization (highly important for N, S, and often P), sorption-des- 
orption (highly important for P and cations) and weathering-precipitation (high- 
ly important for micro nutrients and P). 

In this chapter, results are summarized from earlier and ongoing nutrient 
balance studies in Africa at subcontinental, subnational, farm, and field levels. 
Differences in interpretation among these studies are discussed and future 
avenues for nutrient-balance research given. 


Subcontinental Scale 

In the late 1980s, FAO replaced its fertilizer-driven philosophy by an INM 
approach, which triggered a debate on high versus low external input farming. In 
this context, FAO commissioned a study on nutrient balances in agricultural sys- 
tems in Africa, with the aim of creating awareness on not just the state of soil fer- 
tility in the subcontinent but also its dynamics. The nutrient balance study for 38 
African countries (Stoorvogel & Smaling, 1990; Stoorvogel etal., 1993) involved 
partitioning the continent into rainfed cultivated, irrigated, and fallow land, for 
which FAO provided hectarages and yields. Rainfed land was further divided on 
the basis of the length of the growing period and the soil map of Africa, at a scale 
of 1:5 000 000 (FAO/UNESCO, 1977). The basic spatial unit was the land-use 
system, for which five nutrient inputs and five nutrient outputs were calculated or 
estimated (Table 2-1). For this exercise, many country statistics, maps, reports, 
and literature were scrutinized. A detailed account of the information garnered 
and interpreted is annexed to the main document by Stoorvogel and Smaling 

The amount of data available to calculate the five inputs (IN 1 to 5) and the 
five outputs (OUT 1 to 5; Table 2-1) varied largely between and within countries. 
As a consequence, much available detail had to be dropped and discrete ratings 
developed for variables that normally represent a continuum. Also, average val- 
ues were used for properties that showed wide ranges, such as crop-nutrient con- 
tents. Quantitative information on atmospheric deposition, leaching, and gaseous 
losses was very scarce. Instead of going by educated guesses, transfer functions 
were built (Bouma & Van Lanen, 1987; Wagenet et al., 1991). These are regres- 
sion equations, in which the nutrient flow is explained by parameters that are easy 
to measure. For leaching, for example, the equations represent the best fit for a 
series of point data on leaching, which were accompanied by such building 


Table 2-1. Nutrient inputs and outputs calculated in continental and district studies. 

Nutrient inputs 

INI Mineral fertilizers 

IN2 Organic inputs (manure, feeds, waste) 

IN3 tin ii i hi in i u i I du i 

IN4 Biological nitrogen fixation 

IN5 Sedimentation by irrigation and natural flooding 

Nutrient outputs 

OUT 1 Harvested products 

OUT 2 Crop residue removal 

OUT 3 Solute leaching 

OUT 4 Gaseous losses 

OUT 5 Runoff and erosion 

blocks as rainfall, soil fertility class, and use of fertilizer and manure. Soil fertil- 
ity classes were merely rated low (1), moderate (2), and high (3) on the basis of 
soil taxonomy (sub)orders. Mollisols, for example, were ranked 3, whereas 
Psamments were ranked 1. For erosion, quantitative information on soil loss was 
amply available, but its translation into nutrient losses was seldom studied. 
Moreover, the studies were often done at the miniplot level, the results of which 
cannot be linearly scaled up to the watershed. 

The results can be portrayed per land-use system, per agroecological zone, 
per country, and also per nutrient for the entire continent. The average N, P, and 
K balances for Africa were -22, — 2.5, and -15 kg ha" 1 yr 1 , respectively. 
Nutrients exported in harvested products, in runoff, and in eroding sediments 
were high and caused the balances to be negative. The implication of the figure 
is that on average, soils in Africa must supply 22 kg N ha" 1 each year to balance 
the ledger, leading to a decline of the N stocks. Figure 2-1 shows the results aver- 
aged for each country. There are countries with near-equilibrium nutrient bal- 
ances and those with high nutrient depletion. 

Nutrient depletion is most intense in East Africa, next in coastal West 
Africa and southern Africa, and least intensive in the Sahelian Belt and Central 
Africa (Table 2-2). In East Africa, major faulting and volcanic activity have pro- 
duced red fertile soils derived from basalt that are generally known as Nitisols 
(FAO, 1988) — rhodic groups and subgroups of Alfisols and Oxisols (Soil Survey 
Staff, 1992) — and Vertisols at low landscape positions. High nutrient depletion is 
due to high outputs of nutrients in harvested products and erosion and also in the 
relatively high inherent fertility of the soils. Coastal West African countries are 
dominated by Alfisols of moderate fertility, in both humid forest and moist savan- 
na regions. Southern Africa also is dominated by Alfisols, many of which are 
sandy and of low inherent fertility. Often associated with these are Vertisols 
derived from basalt that are intensively cultivated as dambo gardens. The 
Sahelian belt — from Senegal to Somalia — is characterized by sandy Alfisols and 
Entisols, often of extremely low fertility, and irrigated Vertisols and Entisols adja- 
cent to major rivers. Central Africa is characterized by infertile, acid Ultisols, 
Oxisols, and Entisols in both forested and savanna regions. 


Nutrient depletion (kg ha" 1 yr - 1 ) 

[yT] Low < 10 < 1.7 < 8.3 

[!w^ Moderate 10 to 20 1-7 to 3.5 8.3 to 16.6 

E^ High 

g§H Very high 


3.5 to 6.6 

* 6.6 

16.6 to 33.2 
i- 33.2 


Table 2-2. Estimated nutrient-depletion rat 
in 1983, excluding South Africa.t 

;s in cultivated land i 

n subregio 

ns of sub-Saharan Africa 


Parameter Highlands 

West Southern 
Africa Africa 


Central sub-Saharan 
Africa Africa 

Cultivated land (million ha) 39 
Depletion rate, (kg ha'" yr 1 ) 

63 24 


20 201 

Total depletion imillioi 

0.45 0.15 3.03 

f Source, Stoorvogel and Smaling (1990). 

X Countries included in each region as follows East African Highlands: Burundi. Ethiopia, Kenya, 
Madagascar, Rwanda, Tanzania, and Uganda. Coastal West Africa: Benin, Cameroon, Cote 
d'lvoire. Ghana. Guinea. Liberia, Nigeria, Sierra Leone, and Togo. Southern Africa: Angola, 
Botswana Lesoth Ml I ibiqu il i I m nd Zi i\e Sahelian Belt: 

Burkina Faso, Chad Hie Gambia. Mali Mauritania Niger. Senegal onialia and Sudan. Central 
Africa: Central African Republic, the two < ongos. and Gabon. 

Sub national Scale 

The subcontinental scale and uneven data availability implicitly brought 
about a considerable amount of generalization, simplification, and aggregation. 
As a follow-up, similar studies were done at subnational scales, i.e., in the 2200- 
km 2 subhumid Kisii District in Kenya (Smaling et al., 1993) and in the 12 230- 
km semi-arid region of southern Mali (Van der Pol, 1992). Primary data were 
available on climate, soils, land use, mineral fertilizers, farmyard manure, crop 
yields and residues and their nutrient content, and to a lesser extent on erosion. 
Kisii soils are predominantly well drained, very deep, and rich in nutrients 
(Mollisols, Luvisols), with the exception of P. Mean annual rainfall ranges 
between 1350 and 2050 mm. Major food crops in the district are maize and bean 
(Phaseolus vulgaris L.), often grown in association. Major cash crops include tea 
[Camellia sinensis (L.) Kuntze], coffee (Cojfea arabica L.), and pyrethrum 
[Chrysanthemum cinerariaefolium (Trev.) Bocc.]. Most farm holdings in addition 
comprise small improved pastures for livestock. Less than 5% of the land is left 
fallow during a year. In southern Mali, millet [Pennisetum glaucum (L.) R. Br.; 
20% of arable land], sorghum [Sorghum bicolor (L.) Moench; 17% of arable 
land], and cotton (Gossypium hirsutum L.; 15% of arable land) are the major 
crops of the region. Smaller percentages of maize and groundnut (Arachis 
hypogaea L.) are grown. Approximately 29% of the arable land is left fallow in 

Calculations revealed that annual nutrient depletion in Kisii District was 
112 kg N ha- 1 , 2.5 kg P ha" 1 , and 70 kg K ha- 1 (Table 2-3), whereas in southern 
Mali the values were 25 kg N ha- 1 , kg P ha- 1 , and 20 kg K ha" 1 (Table 2-X). In 


Table 2-3. Nutrient budget in Kisii District, Kenya, t 

Element IN 1 IN3 IN 3 IN 4 OUT L OUT 2 OUT 3 OUT 4 OUT 5 Total 

t Source. Smaling et al. (1993). 

Table 2-4, Nutrient budget in southern Mali.f 

IN 2 IN 3 IN 4 OUT 1 + 2 OUT 3 OUT 4 OUT 5 Total 

:, Van der Pol (1992). 

Kisii, removal of nutrients in the harvested product (OUT 1) was the strongest 
contributor to the negative balance, followed by runoff and erosion and, for N, 
leaching. Use of mineral fertilizers and manure in Mali is much less than in 
Kenya, but crop production is also lower, reflected in lower values of the output 
of aboveground crop parts (OUT 1). Because of lower rainfall and flatter topog- 
raphy, losses from leaching, denitrification, and erosion also were smaller in 

At the crop level, conclusions drawn from the Kisii study revealed that 
pyrethrum is the big nutrient miner (-147 kg N, -24 kg P, -96 kg K ha" 1 yr' 1 ), 
whereas tea has the most favorable nutrient balance (-67 kg N, +6 kg P, -30 kg 
K ha" ' yr" ' ) . Pyrethrum receives little mineral or organic fertilizer, has a high 
nutrient content per unit of harvested product and protects the surface poorly 
against erosion. Tea, however, receives substantial amounts of mineral fertilizer 
and offers good protection to the topsoil. In southern Mali, millet is the big nutri- 
ent miner (-47 kg N, -3 kg P, -37 kg K ha" 1 yr" 1 ), whereas cotton has the most 
favorable nutrient balance (-21 kg N, +7 kg P, -9 kg K ha" ' yr" ' ) . Millet receives 
virtually no mineral or organic fertilizer and has a high nutrient content per unit 
of harvested product as compared with sorghum. Cotton, however, receives sub- 
stantial amounts of fertilizer. 

Farm and Field Scale 

The subcontinental and subnational studies revealed that N and P, on aver- 
age, are moderately to strongly mined. In Kisii District, soils are still rich enough 
to produce high agricultural output. But for how long? And how will farmers be 
told not to go for high crop yields when they can obtain them? Should farmers 


apply N fertilizer when the N balance is as negative as - 1 1 2 kg ha" ? These ques- 
tions were posed by many interested parties after publication of the subcontinen- 
tal and subnational studies, and they triggered the development of a proposal for 
a nutrient monitoring programme (NUTMON) at the farm scale (Smaling & 
Fresco, 1993; Smaling etal., 1996). 

In 1995, a Rockefeller Foundation-sponsored NUTMON pilot project 
started in 26 farms in three agroecologically and ethnically different districts in 
Kenya (Kisii, Kakamega, and Embu). The initial phase included interpretation of 
satellite images and identification of more or less homogenous land-use zones. In 
each zone, rural appraisals were then held, which led to the identification of char- 
acteristic farm types for each land-use zone and the subsequent selection of pilot 
farms. For each farm, an initial inventory was done on household composition, 
farm and field architecture, agricultural activities, and nutrient stocks. This was 
then followed by monthly monitoring of farm management activities related to 
nutrient flows and related economic factors (De Jager et al., 1998b). 

Results so far indicate an average negative N balance of-7 1 kg ha" 1 for the 
three districts (Van den Bosch et al, 1998); however, if one just looks at the flows 
that are managed directly by the farmer (mineral and organic fertilizers, and har- 
vested crops and residues leaving the farm), the annual N balances are positive 
(10, 35, and 46 kg N ha" 1 for Kisii, Kakamega, and Embu Districts, respective- 
ly). Phosphorus and K balances were close to equilibrium. It appeared that input 
through manure derived from communal lands, where animals are grazing during 
daytime, is an important nutrient input at the farm level. The virtual absence of 
these communal lands in Kisii explains the lower N balance value. One major 
methodological constraint was that some flows were actually measured, whereas 
others such as leaching and gaseous losses were estimated. Yet they influence the 
value of the balance very much. 

Relations also have been established between economic performance indi- 
cators, the socioeconomic environment, farm management practices, and nutrient 
balances. It was found that net farm income shows no relation to the nutrient bal- 
ance (De Jager et al., 1998a). A high degree of market orientation, however, cor- 
related well and negatively with the N and K balance. The market-oriented farms 
located in the densely populated areas and characterized by intensive crop and 
livestock activities import nutrients through fertilizers and animal feeds, but the 
amount is insufficient to compensate for the outflow through marketed products, 
leaching, and erosion. Subsistence farms in the less populated areas (drier parts 
of Kakamega and Embu) have a relatively successful strategy to concentrate 
nutrients through grazing of cattle in communal lands. Off-farm income also 
proved very important for households to survive. Without this source of income, 
54% of the farms in the sample would be below what the World Bank considers 
to be the poverty line. The replacement costs of mined nutrients amounted up to 
35% of the average net farm income. 

At the crop and field level, cash crops such as tea and coffee realized high- 
er gross margins and considerably lower nutrient mining levels than the major 
food crops, maize, and beans. Application of sufficient nutrients to food crops 
apparently is not viable in the current economic environment (De Jager et al., 



Comparing Subcontinental and Subnational Scales 

The Kisii District study yielded nutrient loss values of-1 12 kg N and -3 
kg P ha" 1 yr" 1 . In the subcontinental study, the extrapolated nutrient balance for 
Kisii District would have been -75 kg N and -5 kg P ha" 1 yr — 1 . In the latter 
study, all soils would have been in Fertility Class 2 (moderate), characterized by 
1 g N kg" 1 soil and 0.2 g P kg" 1 soil. In reality, however, the soils have a higher 
N content, which could be adequately covered in the district study. Pyrethrum 
turned out to be the major nutrient miner in the district study, but it was not 
included in the supranational study because it lacked importance at mat scale. 
Hence, the differences between the results of the two studies are differences in 

Comparing Subnational and Farm Scales 

In the NUTMON pilot, farm-determined nutrient balances for Kisii were 
-102 kg N, -2 kg P, and -34 kg K ha" 1 yr" 1 , which compare well with the sub- 
national estimates (Van den Bosch et al., 1998). Variation around the mean, how- 
ever, was considerable. Nutrient stocks used in the subnational study were aver- 
age values for land units on a 1:100 000-scale soil map for Kisii District (Smaling 
et al, 1993). The six farms in Kisii District had total N concentration between 1.5 
and 4.6 g kg" 1 soil and total P concentration of 0.9 to 1.3 g kg" 1 soil. 

Comparing Farm and Enterprise Scales 

Nutrient stocks of individual plots within farms and village territories can 
differ considerably. Reasons range from differences in soil texture, land-use and 
fallow history to microclimatic differences. Smallholder farmers exploit 
microvariability, because for each weather condition, there are pieces of land 
where crops perform well (Brouwer et al., 1993). Hence, farm and field hetero- 
geneity is often regarded as an asset by those who are resource poor and risk 
averse, their goal being food security rather than bumper harvests. An example of 
taking advantage of heterogeneity is the use of termite mounds, representing 
spots of relatively high fertility. Another striking example of farm-level variation 
is in the ring management systems in semiarid West Africa, where inner circles 
near the farms and village are much more intensively used and managed than 
outer rings (Prudencio, 1993; Sedogo, 1993). Of the three subsystems shown in 
Table 2-5, the homestead fields represent the plots just around the homestead, 
and receive substantial amounts of nutrients from animal manure and household 
waste. As a consequence, soil productivity in this part of the farm remains at a 
relatively high level. 

In the NUTMON pilot project, it became clear that cash crop and food crop 
plots are treated quite differently as regards nutrient flows (De Jager et al, 
1998b). The role of livestock in the farming system and the amount of manure 
reaching certain plots largely determines within-farm differences i 
stocks and flows (Mohamed-Saleem, 1998; Van den Bosch et al., 1998). 


Table 2-5. Nutrient 
of West Africa. 

stocks of different subsystems it 


upland farm in the Sudan-savanna zone 

Farm subsystem 

pH in H 2 

Organic C 

Total N 

Extractable P 

Exchangeable K 


mmol, kg" 1 



Homestead fields 
> in i Id 
Bush fields 





5- 1 6 

t Source, Sedogo (1993). 


The studies discussed in this chapter contributed to the shortlisting of nutri- 
ent balance as a land-quality indicator in a World Bank initiative to capture the 
current quality of land, the pressures exerted on it, and societal responses (Pieri 
et al., 1995). Does nutrient balance qualify as such? Perhaps, but it may be wise 
to first list the many constraints that may preclude its usefulness and the oppor- 
tunities involved. 


Farmers continue to deplete soil nutrients as long as the land provides them 
sufficient food and cash to make it through the year. In Kisii District, for exam- 
ple, gross nutrient mining was observed, but as soil fertility is still rather high, 
crop production was also rather high. In other words, as long as the soil is able to 
buffer the negative balances before reaching low levels of nutrient availability, 
farmers will not notice changing soil fertility the next year. The nutrient balance 
alone is therefore not sufficient as an indicator of soil productivity. It needs to be 
linked with soil nutrient stocks, either with the total stock or with the stock of 
available nutrients. The latter may be defined as the nutrients that are present in 
the soil solution at the beginning of the growing season or will enter the soil solu- 
tion during the season. 

Not all inputs and outputs are easily measured. Determining inputs by min- 
eral fertilizers may require a quick look at district statistics, and yield estimates 
may just require some ground-truth measurements. For leaching and gaseous 
losses, however, transfer functions are needed, which are made up of different 
parameters with values that are often obtained from secondary sources, and hence 
their values are less reliable than those of nutrients in crop products. 

A nutrient balance value may contain considerable error because it reflects 
the aggregation of five inputs and five outputs. The sheer lack of certain cate- 
gories of primary data in the tropics makes it difficult to put the nutrient balance 
concept into operation. There are no examples of benchmark sites where all 10 
parameters of Table 2-1 have been measured simultaneously over sufficiently 
long periods. 

For the continental, national, and district studies, input data for the nutrient 
balance are mostly derived from subdistrict statistics and are thus already aggre- 



; nutrient flows with the sum 



(IN 1 + IN 2)1 

(OUT 1 + OUT 


(Balance 1 + 2)1 

dotal inputs) 

(total outputs) 

(total balance) 

gated to some extent (soil maps, district statistics, national fertilizer-use statis- 
tics). Moreover, the loss of resolution and of relevant human-induced and natur- 
al spatial variability tends to produce average figures and trends without any 
information on standard deviation. 


For the nutrient balance to become a meaningful land-quality indicator, it 
is necessary to develop a quality index, relating nutrient balance to nutrient stocks 
in one way or another. The concept of stocks and flows is in line with economists' 
style of budgeting and may help in bridging gaps between disciplines. Moreover, 
it will be possible to estimate how long nutrient mining in a given land-use sys- 
tem can continue unabated. 

The most straightforward approach would be the use of nutrient balance 
and nutrient stocks where both include all nutrients, irrespective of their avail- 
ability. The strength of such an index is that it provides information on the long- 
term fate of the land and notjust of the next crop. A disadvantage is that it is not 
directly related to the nutrients that are immediately available and hence not to 
crop growth. Another drawback is the difficulty to assess the values of the nutri- 
ent balance, certainly in low-data environments. An index consisting only of 
flows that can be easily determined would have much more practical meaning. 
These flows are IN 1 and IN 2 (mineral and organic fertilizers), and OUT 1 and 
OUT 2 (removed biomass in harvest and crop residues). The values of these four 
flows are all strongly human influenced and directly reflect the farm households' 
allocation of capital and labor as well as income generation and food security 

Disadvantages are that potentially important flows are ignored. This would 
not matter if the proportions of the various INs and OUTs were little affected by 
the type of agrosystem. It is obvious from Table 2-6 that neither IN 1 + IN 2 nor 
OUT 1 + OUT 2 is a constant portion of the total input or output. Dividing the 
truncated balance by the total balance (last two columns of Table 2-6) may even 
lead to negative values (e.g., P in Kisii). Similar results as in Table 2-6 were 
found for the data by Stoorvogel and Smaling (1990), implying that the use of 
only INs 1 and 2, and OUTs 1 and 2 instead of all INs and OUTs does not offer 
good prospects. Nevertheless, the ratio (IN 1 + IN 2)/(total inputs) presents inter- 
esting information, for it indicates the degree of human involvement in nutrient 
supply. The ratio (OUT 1 + OUT 2)/( total outputs) indicates which fraction of the 
outputs can be seen as useful. 


In the quest for a land-quality indicator for nutrients that is also directly 
related to yield, the appropriate flows of available nutrients must first be identi- 
fied. This is simple for the outputs: OUTs 1 through 4 refer to available nutrients, 
while OUT 5 (erosion) also refers to nutrients that are not immediately available 
because they are present in solid organic and inorganic particles, in addition to 
available nutrients. For the inputs the situation is more complex. The nutrients of 
IN 4 are plant available. Those of IN 3 are directly available as far as wet depo- 
sition is concerned (estimated at 50%), whereas those in dry deposition (the other 
50%) are in an unavailable form. Those of IN 5 are not available whenjust con- 
sidering sedimentation and not run-on. The nutrients of IN 1 and IN 2 are partly 
or entirely in an available form. N and K in chemical fertilizers and K in organic 
fertilizers usually can be considered as available. Water-soluble P fertilizers and 
organic fertilizers have about the same fraction of available P; it is often set at . 1 , 
but it varies between 0.05 and 0.2 depending on soil properties and weather con- 
ditions. The availability of N in organic fertilizers is affected by weather condi- 
tions, length of growth season, and type of manure. An often-used default value 
is 0.4. With these assumptions, the following formulas were applied to estimate 
the balance of INs and OUTS of available nutrients: 

for N: (IN 1 + 0.4 IN 2 + IN 3 + IN 4) - (OUTs 1 to 4) 
for P: (0.1 IN 1 + 0.1 IN 2 + 0.5 IN 3) - (OUTs 1 to 3) 
for K: (IN 1 + IN 2 + 0.5 IN 3) - (OUTs 1 to 3). 

The thus estimated values were compared with total nutrient stocks in the 
soils of Kisii and southern Mali. The resulting values for N, P, and K indicated 
annual losses of the nutrient stocks in the order of magnitude of 1.2% for N and 
0.35% for both P and K. These values, however, are very strongly affected by the 
assumptions made in the calculations of the INs and OUTS. 


The nutrient balance results obtained for the subcontinental study paint a 
rather gloomy picture. Soil fertility is really at stake; however, it is risky to draw 
conclusions from low-resolution, aggregated studies. Generally, the largest unit 
for which soil nutrient balances can be quantified is the field, whereas larger spa- 
tial scales can only be dealt with through generalization and aggregations 
(Stoorvogel & Smaling, 1997). For nutrient balances, aggregation is a very deli- 
cate issue, as the balance itself is made up of at least 10 parameters (Table 2-1), 
which are in some cases outcomes of regression analysis on again more basic 
parameters. Also, a negative balance does not necessarily mean that crop pro- 
duction declines instantly because soils may have a large buffering stock of nutri- 
ents, sufficient to keep production going for many years (Smaling et al, 1996). 

Based on this, we suggest that the subcontinental results should be treated 
as general awareness raisers, i.e., that soil fertility decline in Africa is a threat and 
needs attention, just like nutrient accumulation in some parts of Europe needs 
attention. At the national and subnational levels, results are meant to alert nation- 
al and subnational policy makers and other stakeholders. Research and develop- 


ment efforts can be better targeted, but again the results do not reveal much on 
differences in farmers' management and strategies. This becomes visible only 
during farm-level monitoring activities, as carried out during the NUTMON pilot 
(Van den Bosch et al, 1998). Similar work is going on in several African coun- 
tries, such as Kenya (Shepherd & Soule, 1998), Mali (Defoer et al, 1998), 
Ethiopia (Elias et al., 1998), and Tanzania (Baijukya & De Steenhuijsen Piters, 
1998). In the recent past, different authors (e.g., Prudencio, 1993; Brouwer et al, 
1993; Carter & Murwira, 1995; De Steenhuijsen Piters, 1995) have shown how 
risk -averse farmers in West and southern Africa cherish and exploit spatial vari- 
ation in soil fertility. Analogies in the field of soil and water conservation also are 
plentiful (Tiffen et al., 1994; Rey et al., 1996), and clearly signal a warning to 
those who tend to rely only on averages and smoothness of trends. Survival 
strategies of African farmers are apparently underestimated (Scoones & Toulmin, 

When considering the suitability of the nutrient balance as a land-quality 
indicator for nutrients, no index can be put forward as most obvious. Among the 
possible indices, the ratio of nutrient balance to nutrient stocks may be consid- 
ered as the best one, but it is not easy to determine. The rather easy to determine 
ratio (IN 1 + IN 2)/(total inputs) and the more difficult (OUT 1 + OUT 2)/( total 
outputs) have been worked out to some extent in this chapter but do not seem too 
promising. The observed difficulties make it worthwhile to look further for other 
approaches. Theoretically there are opportunities in chemical soil analysis. The 
number of required data, the variability that is to be expected, and the high costs 
make such alternatives not very attractive. 

Can we still say that soil fertility is at stake in Africa? Yes, because apart 
from the results of the studies presented above, there are a number of on-station 
and on-farm medium- and long-term trials that quantitatively support that state- 
ment. Figure 2-2 shows declining soil fertility in a long-term trial in central 
Kenya, and Table 2-7 summarizes changes in soil nutrients observed during the 

Fig. 2-2. Effect ot h i u in n u i 1 i I 10 t ha- 1 yr "''), mineral fertilizer (N-P, 

120 kg N ha 1 yr 1 and 52 kg P ha- 1 yr "'), and FYM (10 t ha- yr- I - N-P fertilizer (120 kg N ha- 1 
yr 1 and 52 kg P ha ' yr" 1 ) on soil organic C at t-o 25-cm depth at Kabete, Kenya (S.M. Nandwa, 
1997, unpublished data). 


Table 2-7. Changes 
Kenyan soils, + 

in soil properties i 

ii the top 20-cm layer i 

ii nonfertiiized. continuously c 



Organic C 

Tl T2 

Mehlich P 
Tl T2 

Exchangeable K pH i 
Tl T2 Tl 

11 UNO 


Alfiso! (clayey)! 
Alfisol (sandy )§ 

Ultisol (clayey)t 
UltisoL (claycy)j 
Ultisol (loamy)! 
Ultisol (sandy )tt 



t Source, Smaling and Braun (1996). 
t Tl = 1988, T2 = 1991. 
§ Tl = 1988. T2= 1990. 
1 Tl =1988. T2 = 1992. 
I Tl =1987, T2= 1990. 
tt Tl = 1987, T2 = 1991. 

Fertilizer Use Recommendation Project in different parts of Kenya (Smaling & 
Braun, 1996). But is soil fertility at stake all over Africa? No, certainly not! The 
average nutrient balance may be negative, but thousands of farms will be able to 
show sustainable nutrient management strategies at satisfactory production lev- 
els. And if researchers, farmers, and other stakeholders in the agricultural sector 
are ready to learn, listen, and subsequently teach, we may be on our way to a bet- 
ter future for agriculture in Africa. 


Baijukya, F.P., and B. De Steenhuijsen Piters. 1998. Nutrient balances ami their consequences in the 

ana ii , -I i tl incl ii /stem f Bukoba Dish orthu t Tanzania. /// E.M. A i Dialing 

(ed.) mi nl II I il i ii Ii i i i 1 id tainal ilit ii il 

Saharan African a tj o-i isystem gri I i yst Environ, (in press). 

Eiouma. .1.. ami H.A..I. Van Lanen. 19X7. Transt'ei functions and threshold values: From soil charac- 
i i i ii l i i i i p. II 110 k et I.) Quantifi ml evaluation pi 

i ii 1 1 ' in I Int. Inst i ] i I Earth r n i I ill i il I 

I i l I II I t I I ! I 

African semi-arid tropics: A possible risk-reducing factor for subsistence farmers. Agile. 
Ecosyst. Environ. 45:229—238. 

Carter. [ md U Vlurwii 199 itial iriabilit) n il i rtilit m nagement and crop 

response in Mutoko Communal Area. Zimbabwe. Ambio 24:77-84. 

Defoer, T., H. de Groote, T. Hilhorst, S. Kant6, and A. Budelman. 1998. Participatory action research 
I I it I i i l I I 1 111 it 111 11 111 11 I 111 I i I l in itful m I n ' 

E.M. A. Smaling (ed.i Nutrient flows anil balances as indicators of productivity and sustain- 
ability in sub Sail ran African agi i in i t. 1 nviron (in press). 

De Jager, A., I. Kariuki. F.M. Matiri. M. Odendo, and J.M. Wanyama. 1998a. Monitoring nutrient 
flows and economic performan i farmii tern UTMO 1 Monitorin i 

farm economic performance in three districts in Kenya. In E.M. A. Smaling (ed.i Nutrient 
flows and balances as indicators of productivit) and sustainabilit) in sub-Saharan African 
tgi 1 'i i , i i I ii ii (in press). 

De Jager, A., Nandwa, S.M., and P.F. Okoth. 1998b. Monitoring nutrient flows and economic perfor- 
mance in African farming systems (NUTMON). 1. ( onceptual framework. /// E.M. A. Smaling 


(ed.l Nutrient flows ami balances as indicators of productivity and sustainability in sub- 
Saharan African agro- ystem igri Ei \st Environ. (In press). 

D i iilm i u I ii 1 19 I 11 i n i 11 I i nd farm ' I i ning Meld variations in 

northern i Ph.D. thesi ningen Agri Uni i nil n th i land 

Elias, E., S. Morse, and D.G.R. Belshaw. 1998. The nitrogen and phosphorus balances of some Kindo 
Koisha farms in southern Ethiopia. /// E.M.A. Smaling led.) Nutrient flows ami balances as 
indicators of productivity and sustainability in sub-Saharan African agro-ecosystems. Agile. 
Ecosyst. Environ, (in press). 

FAO. 1988. Revised legend of the FAO-UNESCO soil map of the world. Food and Agriculture 
Organization of the United Nations, Rome. 

FAO/UNESCO. 1977. FAO-UNESCO soil map of the world, 1: 5 000 000. Vol. 6. Africa. United 
Nations Educ. Scientific and Cultural Organ., Paris. 

VI chained I m i ' 199 lutrient balain pattern in tf'ri n ir tock /stem In 1 •! j 

maling led.) lutrient fit mil balanct i indicato f pi luctivil id tai lit in 

sub-Saharan African agro-ecosystems. Agile. Ecosyst. Environ, (in press!. 

Pieri, C. 1989. Fertilite des terres de savanes. Ministere de la Cooperation et CIRAD-IRAT, 
Montpellier, France. 

Pieri, C, J. Dumanski, A. Hamblin, and A. Young. 1995. Land quality indicators. World Bank Dis. 
Pap. 315. World Bank, Washington, DC. 

Prudencio ', 1993. Rim management o1 >il mi rop in ill. ' I \frican emi-arid ti pii Hie 
case of the mossi farming system m Burkina Faso. Agric. Ecosyst. Environ. 47:237-264. 

Rev. C. 1. Scoones. and ('. Toulmin. 1946. .Sustaining the soil. Indigenous ..oil and water conserva- 
tion in Attn. l iu in!' ibl London. 

Scoones. 1., and C. Toulmin. 1908. Soil nutrient balance:,: What use lor policy? In E.M.A. Smaling 
(ed.l Nutrient flows and balances as indicators of productivip and sustainability in sub- 
iharan \fricai igro-ei ■ > tem \grii -.. \ I Environ, (in press). 

Sedogo, P.M. 1993 I luti id i| n inei I i u ultu 'u d ,,, e des modes de ges- 

tion sur la fertilite. Ph.D. thesis. Univ. of Abidjan, Cote d'lvoire. 

Shepherd, K.D., and M.J. Soule. 1998. Soil fertilitj management in West Kenya: Dynamic simulation 
of productivity, profitability and sustainability al different resource endowment levels. In 
E.M.A. Smaling (ed.l Nutrient flows and balances as indicators of productivity and sustain- 
ability in sub-Sah m frit n agi ten gric. Ecosyst Environ Impress). 

Smaling, E.M.A., and A.R. Braun. 1996 .,1 i rtilitj u hut ub- iha tn Attica New dimen- 

sions, new challenges. Commun. Soil Sci. Plant Anal. 27:365-386. 

Smaling, E.M A . an I 1 O. Fn op i dc ision-su port nun I for monitoring nutrient balances 

under agricultural land use (NUTMON). Geoderma 60:235-256. 

Smaling, E.M.A., L.O. Fresco, and A. de Jager. 1996. Classifying, monitoring and improving soil 
nutrient stocks and flows in African agriculture. Ambio 25:492-496. 

Smaling, E.M.A I 1 i i I. ami P v'indmeij 1993. Calculatin 1 nutrient balances in 

Africa at different scales: 11. District scale. Fert. Res. 35:237-250. 

Soil Survey Staff. 1992. Keys to soil taxonomy. 5th ed. Soil Man _ i, pc ii ei\ Tech. Monogr. 
19. Pocahontas Press, Blacksburg, VA. 
i 1 I i i I I I U n in ub Sahatan 

Africa. 1983-2000. Rep. 28. DLO Winand Staring Ctr. for Integrated Land, Soil and Water 
Res., Wageningen, the Netherlands. 

Stoorvogel, J.J., and E.M.A. Smaling. 1997. Research on soil fertility decline in tropical environ- 
ui ii In ti ii i L a l mi 'hi t i (in press). 

Stoorvogel, J.J., E.M.A. Smaling, and B.H. Janssem 1993. Calculating soil nutrient balances at dif- 
ferent scale I upi lational il 1 i i 
fiffen I I 1 urn i n I i i id i 1 I i frl u 1 n u onmental recovery 

in Kenya. John Wiley & Sons, New York. 

Van den Bosch, H., S. Maobe, V.N. Ogaro, J.N. Gitari, and J. Vlaming. 1998. Monitoring nutrient 
I i ii in i n Mi n t i in mi i UTMO 111 I i in i 

i mil m Hi in th n ii i in u ' L M i in ilin ( i 1 urn n flo i 1 il in 

as nidi ,ii i hi p ic ion md ii i an ilnlio in i i ihai u '1 can gro-eci i in 
Agric. Ecosyst. Environ, (in press). 
in il i P 1 i 1992 1 mining ii mi ii i iuil utorto 1 n in n in < ith ,n M Ii 1 till 
Royal Tropical Inst.. Amsterdam, the Netherlands. 

Wagenet, R.J., J. Bouma, and R.B. Grossman. 199 I . Minimum data sets for use of soil survey infor- 
mation in soil interpretive models, p. 161-182. In M.J. Mausbach and L.P. Wilding led.; 
Spatial variabilities of soils and landforms. SSSA Spec. Publ. 28. SSSA, Madison, WI. 


Soil Fertility Management in 
Africa: A Review of Selected 
Research Trials 

Mateete A. Bekunda 

Makerere University 

Kampala, Uganda 

Andre Bationo 

International Fertilizer Development Center 
Niamey, Niger 

Henry Ssali 

/ ' • ' ' 'i i •;/ 

Kampala, Uganda 


The increasing recognition of soil fertility depletion as the main biophysical factor 
limiting crop production in many African smallholder farms has raised interest in using 
data from past fertilizer studies to identify options for increasing agricultural production. 
This review of selected fertility research trials in sub-Saharan Africa reveals a pool of 
information (i) on the principles of fertilizer application for efficient nutrient use and (ii) 
on potential problems arising with continuous use of fertilizers in intensively cultivated 
systems. Adequate soil fertility for sustained crop yields can be obtained with combined 
use of mineral fertilizers and organic materials. Continuous use of N fertilizers can acidi- 
fy soil, which then requires liming when organic inputs are limiting. Increased deficien- 
cies ofN, P, and other nutrients can be expected as a result of intensive cultivation and 
unbalanced fertilizer use. The use of mineral fertilizers by many smallholder farmers 
remains low because of socioeconomic constraints. This suggests that locally available 
organic materials will continue to be used as sources of nutrients. Future soil fertility tri- 
ll houlcl. therefore, partial larl tint at identifying praetict for judicious u >f organi 
materials and theii combination with mineral fertilizers. Shortcomings of past soil fertili- 
ty research include limited economic analysis of results and use of trial sites and manage- 
ment that poorly represented those of smallholder tanners. Future research should strive 
for active participation of fanners, longer time frames to fully evaluate residual effects and 
rigorous economic analysis of results. 


Sub-Saharan Africa (hereafter referred to as Africa) is characterized by diverse 
agricultural systems that are typically low input and based on subsistence farm- 
ing. Traditional shifting cultivation and rotational systems permitted low but rel- 
atively stable food production on relatively poor soils. Rotational fallows are eco- 
logically sound in low capital input areas (Sanchez, 1976). As population and 
pressures for land use increase, fallow intervals decrease in length until perma- 
nent intensive cultivation predominates. With time, these farms become subdi- 
vided and more intensively cultivated. Shifting cultivation continues to be prac- 
ticed in only isolated locations and often at the expense of dwindling forest 
reserves and other natural habitats (Tukahirwa, 1992). 

Substantial quantities of nutrients can be lost in permanent cultivation sys- 
tems through offtake in harvested products and crop residues and through loss by 
runoff and leaching and as gases. Stoorvogel and Smaling (1990) and Stoorvogel 
et al. (1993) estimated high net nutrient losses totaling about 9.3 million tonnes 
in 1983 in Africa. Nutrient depletion can be particularly high in countries with 
high population densities, such as Ethiopia, Kenya, Malawi, and Rwanda 
(Smaling etal, 1997, this publication). Even more alarming is the projected fig- 
ure for nutrient losses of up to 13.2 million tonnes in Africa by the Year 2000. 

Nutrient depletion is complicated by the low inherent fertility of many 
soils, of which >80% have chemical or physical limitations to crop production 
(Sanchez, 1976). Nutrient depletion has been suggested as the main biophysical 
factor contributing to decreasing agriculture production in Africa since the mid- 
1960s (Vlek, 1993; Van Reuler & Prins, 1993; Sanchez et al., 1997b, this publi- 
cation). Replenishment of soil fertility has therefore been singled out as one main 
activity likely to result in positive benefits associated with enhanced crop pro- 
duction, increased coverage of the soil surface with vegetation, and increased soil 
biological activity (Sanchez et al., 1997a). 

Soil fertility replenishment can be achieved through the use of fertilizers, 
both organic and mineral. Where fertility depletion is already high, relatively 
small amounts of crop residues and animal manures are produced, and mineral 
fertilizers become the principal sources for building up nutrients in soils. But 
before fertilizer use is adopted, farmers must be aware of the fertilizer forms, 
methods of use, and potential benefits accruing from their use. In this chapter, we 
review selected soil fertility trials in Africa in terms of their contribution to the 
pool of information on increased and sustainable yields from efficient fertilizer 


The semiarid zone has been defined by Deckers (1993) as an area with a 
growing period ranging from 75 to 179 d. Entisols, Alfisols, and Vertisols are the 
main soils of the area. Entisols are mainly composed of quartz and have low 
water-holding capacity and nutrient content. Entisols are weakly structured and 
prone to water and wind erosion. Leaching can hamper efficient use of fertilizers 
on Entisols. Alfisols have a clay accumulation horizon and a low capacity to store 
plant nutrients. Vertisols are characterized by a high content of swelling clay with 


usually high fertility, except that P availability is generally low and high N loss- 
es can occur under waterlogging conditions. 

Soils in the semiarid zone generally have low organic C and total N con- 
tents because of low biomass production and a high rate of decomposition 
(Mokwunye et al, 1996). Nitrogen and P are limiting nutrients. Soil P stocks are 
low, but the low-activity clay of these soils has a relatively low capacity to fix 
added P (Bationo & Mokwunye, 1991). Therefore, the P requirement for maxi- 
mum yield is often low (Mokwunye, 1979; Osiname, 1979). 

Characteristics of the soils in the semiarid zones present problems of effi- 
cient use of applied N. Studies with 15 N in the semiarid tropics (Ganry et al, 
1978; Gigou & Dubernard, 1979; Chabalier & Pichot, 1979) indicate increased 
loss of N from applied fertilizer with increasing rates of fertilizer application and 
high N losses regardless of N sources. Mughogho et al. (1986), in a summary of 
15 N research by the International Fertilizer Development Center (IFDC) in semi- 
arid areas of West Africa, reported that calcium ammonium nitrate (CAN) sig- 
nificantly outperformed urea in plant N uptake, which was translated into signif- 
icantly higher yields of pearl millet [Pennisetum glaucum (L.) R. Br.]. Total plant 
uptake of fertilizer N, however, was low (20 to 37%), and losses were severe (25 
to 53%). The majority of N remaining in the soil was found in the 0- to 15-cm 
layer. Ammonia volatilization was believed to be a cause for the N loss. 

Field studies to compare N sources and methods of placement 
(Christianson & Vlek, 1991) showed that millet uptake of l5 N was almost three 
times higher from point-placed CAN than point-placed urea. Fertilizer N uptake 
by the plant was reduced by 57% when CAN was broadcast rather than point 
placed. Although urea acidifies the soil faster than CAN, its higher analysis 
appears to outweigh any deleterious effect induced in soil by its use. The high 
soil-acidifying effect of ammonium sulfate has made it less popular. 

Split application of N fertilizer can considerably improve the efficiency of 
the applied N (Uyovbisere & Lombin, 1991). In a l5 N experiment conducted in 
southern Niger with sorghum [Sorghum bicolor (L.) Moench], 8% of the applied 
N for the first split and 19% for the second split were recovered in the grain 
(ICRISAT, 1989). Large positive and additive effects of crop residue and fertiliz- 
er application have been reported in the Sahelian zone (Fig. 3-1), where soil 
organic matter (SOM) content is low. Grain yields in the control plots were low 
and steadily declined during this 4-yr study. The results indicate that continuous 
cereal grain production on these soils will be more successful when mineral and 
organic fertilizers are combined (Janssen, 1993; Palm et al, 1997, this publica- 
tion). In the Sudanian zone, long-term studies indicate that sustainable sorghum 
production can be obtained only when mineral fertilizers are combined with 
manure (Fig. 3-2). The combination of organic materials with mineral fertilizers 
also improves SOM and pH (Bationo et al., 1995). 

Although yield responses to fertilizer and organic inputs are generally pos- 
itive in experiments in the semiarid zone, the responses can vary with the amount 
of rainfall (Pieri, 1973; IRAT, 1974). Split applications of N fertilizer can be 
adjusted during the season according to the degree of water stress (Piha, 1993), 
and conservation of water can enhance the beneficial effects of fertilizer applica- 
tion (Mokwunye et al., 1996). In farmer-managed trials in Burkina Faso, sorghum 


1M6 1986 

pearl millet gram yield in Niger (adapt- 

grain yields were higher with the combination of fertilizer and tied ridges than 
with either fertilizer or tied ridges alone (Nagy et al., 1990). In Zimbabwe, 
sorghum yields were increased from 118 to 388 kg ha- 1 using 15 m tied ridges, 
and to 1071 kg ha' 1 when 50 kg N ha" 1 was applied to the tied ridges during a 
low rainfall season (Nyakatawa, 1996). Thus, for the 34% of Africa that is semi- 
arid and characterized by unreliable and low rainfall, the relationship between 
soil water balance and crop yields plays a major role in the use of fertilizers. 

Cropping systems have been observed to influence N use efficiency. 
Bationo et al. (1997, unpublished data) found that mean grain yields for 4 yr were 
lower for continuous cropping of pearl millet at N rates from to 45 kg N ha" 1 
than for millet-cowpea [Vigna unguiculata (L.) Walp. sp. unguiculata] and mil- 
let-groundnut (Arachis hypogea L.) rotations. Higher responses with rotations 
than with cereal monoculture have similarly been obtained for a maize (Zea mays 
L.)-cowpea rotation in Zimbabwe (Mukurumbira, 1985). In Malawi, MacColl 
(1989) showed that grain yield of the first crop of maize following pigeonpea 

n the Sudanian zi 


[Cajanus cajan (L.) Millsp.] averaged 2.8 t ha" higher than that following con- 
tinuous maize with 35 kg N ha" 1 each year. 

Considerable P research in the semiarid zone of West Africa has focused on 
the suitability of direct application of PR as an alternative to soluble P fertilizers 
(Gerner & Mokwunye, 1995). Direct application of ground, reactive PR (i) 
redresses P deficiency, (ii) has a strong residual effect, and (iii) does not acidify 
the soil. The agronomic effectiveness of PRs depends on their chemical and min- 
eralogical composition and soil and plant factors. West African PRs tested in field 
trials include Tahoua and Pare W from Niger, Tilemsi from Mali, Kodjari from 
Burkina Faso, and Hahotoe from Togo (Sedogo et al., 1991; Bationo et al., 1992, 
1997). The results indicate that Tilemsi PR and Tahoua PR could be viable alter- 
natives to soluble imported fertilizers. Research by IFDC and collaborating 
national institutions has shown that partial acidulation of low-reactivity PRs, such 
as Pare W PR, results in improved performance (Bationo et al., 1986; 1992; 
Buresh et al, 1997, this publication). 

Low soil fertility and low use of organic and mineral fertilizers are the 
greatest biophysical constraints to increasing agricultural productivity in farming 
systems in the semiarid region of Africa. The form, method, and timing of appli- 
cations of the limited nutrient sources are important for efficient nutrient use for 
given water regimes and farming systems. Soil fertility in intensified farming in 
the semiarid zone can be maintained only through (i) the efficient recycling of 
organic materials, such as crop residue and manure, in combination with mineral 
fertilizers and (ii) the use of rotations with legumes (Giller et al., 1997, this pub- 


Deckers (1993) has defined the subhumid and humid zones as regions with 
a growing period of 180 to 269 d and >270 d, respectively. The predominant soils 
of these zones are Alfisols, Ultisols, and Oxisols. Alfisols are common in the sub- 
humid tropics. They are frequently deficient in N and P, and they tend to acidify 
under continuous cultivation. Ultisols and Oxisols are well drained, contain little 
or no weatherable minerals, and have a clay fraction containing kaolinite and 
oxides and hydroxides of Fe and Al. These soils typically have low cation- 
exchange capacity and low inherent fertility. They frequently require balanced 
fertilization with several nutrients. Phosphorus sorption is associated with 
hydrous oxides of Fe and Al (Juo, 1981; Le Mare, 1981), and fertilizer P require- 
ments tend to follow the order Oxisols > Ultisols > Alfisols (Warren, 1992). 


Numerous field experiments have demonstrated crop responses to small or 
moderate amounts of P fertilizers and residual benefits of P fertilizers to crops in 
seasons following the P application (Le Mare, 1959, 1974; Boswinkle, 1961). 
Jama etal. (1997) found that broadcast application of 10 kg P ha' 1 as triple super- 
phosphate (TSP) to maize on acid soils in western Kenya (Kandiudalfic 


Eutradoxs and Kandiudalfs, depending on the topsoil clay content) had a signif- 
icant residual benefit to maize in the season following P application. Phosphorus 
fertilization at the tested 10 and 30 kg P ha" 1 rates was financially attractive for 

Early literature highlighted the importance of mineralization of SOM as a 
source of plant-available P. Foster (1976) established that responses of cotton 
(Gossypium hirsutum L.), groundnut, and finger millet [Eleusine coracana (L.) 
Gaertn.] to P fertilizers on ferralitic soils in Uganda were inversely correlated to 
the level of SOM. A strong direct relationship between total soil organic P (P ) 
and plant-available P was observed in Kenya (Friend & Birch, 1960) and south- 
ern Nigeria (Adepetu & Corey, 1976). The ability of SOM to supply P, however, 
has presumably decreased with the decline in total SOM as traditional cropping 
systems containing fallows were replaced by continuous cropping with little or 
no inputs of nutrients. Maroko et al. (ICRAF, 1997, personal communication) 
observed severe P deficiency on a Kandiudalfic Eutrudox continuously cropped 
with maize in western Kenya, even though the soil contained 0.30 g Po kg- 1 soil 
(68% of the total soil P). Rotation of maize with a sesbania [Sesbania sesban (L.) 
Merr.] fallow significantly increased P in microbial biomass and light-fraction 
SOM and slightly reduced P deficiency for subsequent maize crops (Buresh & 
Tian, 1997; Buresh et al., 1997, this publication). The sesbania fallow, however, 
did not eliminate P deficiency, and P fertilization of maize was necessary for the 
sesbania-maize rotation to be financially attractive (ICRAF, 1997). 

Early research examined the management of P to minimize contact 
between fertilizer P and soil, such as through banding, in P-fixing soils. Fox and 
Kang (1978) showed that banding of P fertilizer for maize was beneficial only at 
suboptimal rates (8 or 16 kg P ha" 1 ) on a sandy Alfisol in Nigeria. At P rates to 
obtain the maximum yield of 3.75 t ha" 1 of maize grain, it was better to incorpo- 
rate the fertilizer in the full volume of the soil. Banding may be satisfactory in 
soils with moderate P-sorption capacity, but in high P-fixing soils the band can 
limit root development and subject the crop to other nutrient stresses or water 
stress (Sanchez, 1976). An initial broadcast application of P can be required on 
soils with high P-sorption capacity and low available P (Yost et al, 1979). 

With perennials, the uptake of P from within the soil profile and hence the 
optimal placement of P fertilizer can vary between dry and wet seasons. For 
banana (Musa sp.) in Uganda, Ssali (1972) observed that uniform surface (0 to 15 
cm) placement of P fertilizers was effective in the wet season. In the dry season, 
root activity in the topsoil was much lower suggesting that deeper fertilizer place- 
ment or irrigation after fertilization was required in the dry season. Research with 
coffee (Cojfea arabica L.) in Kenya revealed that root activity and uptake of P 
was highest near the soil surface in the wet season, but in the dry season when the 
topsoil dried out, the root activity was highest at 45 to 75 cm depth (IAEA, 1975). 

A correction and maintenance fertilization approach with a one-time high 
rate of P fertilizer to reestablish optimum soil fertility levels for high productivi- 
ty followed by periodic maintenance applications of P fertilizer has been pro- 
posed (Pieri, 1987; Sanchez et al., 1997b, this publication). This approach may 
be particularly attractive with medium- or high-reactive PRs applied to acid soils 
near the PR source. The added PR would provide a gradual release of plant-avail- 


able P and residual benefit for several years (Rajan et al, 1996; Sanchez et al, 
1997b, this publication). 


When SOM in the topsoil is above 30 g kg" 1 soil, as is the case on opening 
some subhumid and humid lands for cultivation, little or no response to N fertil- 
izers may be obtained (Sobulo & Osiname, 1986; Mughogho et al., 1990). The 
supply of plant-available N from SOM diminishes within a few years because of 
the fast breakdown of SOM and the inability of the soil to retain the released N 
(Giller et al., 1997, this publication). Thus, many field experiments in the subhu- 
mid and humid zones have shown response of nonleguminous crops to N fertil- 
izer (Jones etal, 1960; Scaife, 1968; Christianson & Vlek, 1991). Christianson 
and Vlek (1991), at a range of sites in West Africa, found (i) comparable maize 
yields with banded and point-placed N fertilizer, (ii) superiority of banding and 
point placement compared with broadcast application of N fertilizer, and (iii) 
comparable effectiveness of urea and CAN as N sources. Ssali (1990) similarly 
found statistically comparable yields of maize and common bean [Phaseolus vul- 
garis L.] following application of either CAN or urea on an Oxic Paleustult in 
Kenya. They concluded, however, that high-analysis urea (46% N) was finan- 
cially more attractive than CAN (26% N) because of lower transport costs. Split 
application of N fertilizers to synchronize N supply with plant uptake of N is crit- 
ical for high N use efficiency and minimal N loss (Arora & Juo, 1982; Mughogho 
etal, 1990). 

Continuous use of N fertilizers, especially ammonium sulfate, induces soil 
acidity (Stephens, 1969; Wapakala, 1976; Aduayi, 1984; Juo et al., 1995). 
Stephens (1969) found at nine stations in southern and western Uganda that soil 
pH at to 20 cm was reduced by about 0.3 units per 2.5 t fertilizer ha- 1 after 
application of ammonium sulfate 4 yr. On a kaolinitic Alfisol at Nigeria, pH 
decreased from 5.8 to 4.5 during 5 yr of continuous maize cropping widi ammo- 
nium sulfate fertilizer (Juo etal., 1995). 

One way of minimizing soil acidity from fertilizer N is by supplying N 
through biological Ni fixation. Biological Ni fixation from legumes can sustain 
tropical agriculture at moderate levels of output (Giller et al., 1994, 1997, this 
publication). Under favorable conditions, green manure crops can accumulate 

100 to 200 kg N ha- 1 in 100 to 150 d in the topics (Giller et al., 1994). Rhizobial 
inoculation in East and Central Africa can enhance yield of exotic food legumes 
with specialized rhizobial requirements {e.g., soybean [Glycine max (L.) Merr.] 
from China, pea (Pisum sativum L.) from Asia Minor, and common bean from 
Central America; Table 3-1 }. Phosphorus fertilization can be necessary for effec- 
tive growth and N2 fixation by legumes (Ssali & Keya, 1986; Cassman et al., 


Other Nutrients 

Nutrients other than N and P have received relatively little attention, and 
reports on their deficiencies are limited (Le Mare, 1984; Kang & Osiname, 1985). 


Table 3-1. Rhizobial inoculation response h\ legumes in East Africa. t 






Common bean 










Homa Bay 







2.] 5 


+Source, Woomer et al. (1998). 

Reasons for limited attention to nutrients other than N and P are that (i) traditional 
practices of fallowing are able to increase availability of nutrients in the topsoil 
(Jaiyebo & Moore, 1964; Stephens, 1967), (ii) the supply of these nutrients from 
soil is frequently sufficient for most crops for a number of years before response 
occurs (Foster, 1979), (iii) some N and P mineral fertilizers contain additional 
nutrients (e.g., sulfur in single superphosphate and ammonium sulfate), and (iv) 
organic inputs and crop residues contain nutrients in addition to N and P. 

Anderson (1973) in a review on K responses of various crops in East Africa 
reported that responses to K fertilizers tended to increase with time following 
land opening or grass fallows. Singh and Goma (1995) likewise reported 
response of maize to K on an Oxisol in Zambia 3 yr after start of a long-term trial. 
Results with groundnut in Kenya showed some cases of greater response to K 
when it was applied with P, but the response was less striking in the subhumid 
and humid zones than in the semiarid zones (Tag et al., 1972). Research results 
indicate that increased intensity of cropping will lead to greater need for K inputs, 
especially for crops with high offtake of K in harvested products such as some 
root crops and bananas. 

There are increasing reports that application of K or S in combination with 
N and P increases crop yields, suggesting an increased need for inputs of these 
nutrients as N and P deficiencies are alleviated (Vlek, 1990). The application of 
high-analysis fertilizers (e.g., urea and TSP) without S, can with continuous crop- 
ping, lead to S deficiencies (Friesen, 1991). Micronutrient analysis of common 
bean seed collected from Tanzania, Zambia, and Malawi suggested that Mo and 
Cu were the micronutients most likely to limit N2 fixation and bean growth in 
eastern and southern Africa (Brodrick et al, 1995). 

Liming can be essential for high crop yields on soils with high Al satura- 
tion. In Uganda, Foster (1976) concluded that liming was important at pH < 5.25. 
Acid-tolerant crops, such as cassava jManihot esculenta Crantz) and yam 
[Dioscorea esculenta (Lour.) Burkill], may grow well at lower pH values and 
require less lime. Results from a 6-yr study on an acid Typic Paleudult (pH in 
water = 4.6, sand = 67% at to 15 cm) in Nigeria indicate that relatively low rates 


of lime can sustain yields in a maize-cowpea rotational cropping system (Friesen 
et al., 1982). About 90% of maximum maize yield was obtained at 35% Al satu- 
ration. The critical level for Al saturation for cowpea ranged from 25 to 55%, 
depending upon cowpea variety and rate of mineral fertilizer. Liming can 
enhance use of applied fertilizer (Yamoah et al, 1992) and substantially increase 
exchangeable Ca, pH, extractable P, and effective cation-exchange capacity 
(Pieri, 1987; Yamoah etal., 1992; Lunguetal, 1993). 


Table 3-2 summarizes information from select experiments with arable 
crops for at least 7-yr duration. The experiments were basically designed to deter- 
mine the effects of mineral fertilizers and organic inputs on crop yields and soil 
fertility. The results allow an examination of some issues associated with sus- 
tainability of crop production. 

At all sites, there were positive yield responses to one or more nutrients 
added as mineral fertilizers. The responses were consistent for the duration of the 
experiments. This highlights the effectiveness of mineral fertilizers in increasing 
yield in arable farming systems in Africa. This potential is recognized by large- 
scale farmers, who have been able to sustain relatively high yields of maize 
(Kenya, Zambia, and Zimbabwe), tobacco (Nicotiana tabacum L.; Malawi and 
Zimbabwe), and coffee (Coffea arabica L.; Kenya) for periods of up to 30 yr. 

In 9 out of 13 sites shown in Table 3-2, yields for mineral fertilizer treat- 
ments declined during the experiments. A decline in crop yields with application 
of only mineral fertilizer is further illustrated with results for maize during 19 
years in Kenya (Fig. 3-3). Maize yields for application of mineral fertilizer start- 
ed to become less than for farmyard manure (FYM) and mineral fertilizer plus 
FYM plus crop residues (CR) after 10 yr. Such declines might result from (i) soil 
acidification by the fertilizers, (ii) mining of nutrients as higher grain and straw 
yields remove more nutrients than were added (Scaife, 1971), (iii) increased loss 
of nutrients through leaching as a result of the downward flux of nitrate when fer- 
tilizer N is added, and (iv) decline of SOM. The depletion of nutrients, not added 
with fertilizers, to deficiency levels in the soil is a plausible explanation for 
declining yields in many cases. Many studies have shown that after replenishing 
nutrients initially limiting crop production, other nutrients quite often start to 
become limiting (Kenya Agricultural Research Institute, 1994; Singh & Goma, 

Pieri (1995) reported that use of N fertilizer with sorghum monocropping 
in Burkina Faso accelerated the annual rate of SOM loss from 1.5% without fer- 
tilizer, to 1.9% with moderate rates of N fertilizer, and 2.6% with high N rates. In 
Kenya (Fig. 3-3), however, the application of NP fertilizer did not increase the 
decline in soil organic C. The decline in soil C over 16 yr was comparable for no 
fertilization (20.4-11.8 g C kg" 1 soil), application of NP mineral fertilizer 
(19.5-12.2 g C kg- 1 soil), and application of only FYM (19.6-13.0 g C kg- 1 soil); 
see Fig. 2-2, Smaling etal, 1997, this publication). Soil organic C after 16 yr was 

Table 3-2 

Impact of soil management tre 

atments on crop yield trends in selected Ion 







Test crops 


Cdte D'lvoire 



1969 to 1990 




1969 to 1990 





1957 to 1974 




1957 to 1974 



Burkina Faso 



1962 to 1984 

Groundnut, millet 

1960 to 1983 

Groundnut, millet 







1981 to 1988 







1966 to 1981 

Maize, groundnut 


Lll\ isol 

1966 to 1981 

Maize, groundnut 



1966 to 1981 

Maize, groundnut 





1937 to 1964 

Cotton, millet, sor 






1964 to 1975 

Cotton, millet, sor 





1976 to 1996 

Maize, bean 

+ . yield higher than control; ++, yield 

relatively higher than 

+ within the row; S, s 

t Sources: 1, Traore & Harris (1995); 2 

3, 4, Pieri (1995); Laryea et al. (1995); 5 

Byalebeka (1996, personal communica 

on); McWalter and Wimble, (1976); 8, Sinf. 

Appro \ 

rmate USDA Soil 

Taxonomy ec 

uivalents: Ferralitic, 

Oxisol; Ferruginous, A 

b-Saharan Africa. 


i bl i Id ti dm L» m i in ible v i I I .1 i lin [>I> n ii ield de< liii 
'i, Singh iV i 1 < i I i r i i i i i 1 1 i I 

and Balasubramanian (1979); 9 = Swift et al. (1994) and S. Nandwa (1996, f 

isol; and Lux isol. Alfisol. 


slightly higher (15.6 g C kg" soil) with combined application of NP mineral fer- 
tilizer and FYM. 

The application of organic inputs as either animal manures or crop residues 
increased yields, but in many cases yield tended to decline with application of 
only organic inputs (Table 3-2). Crop residues were generally less effective than 
animal manures as a source of nutrients. Application of residues with high C-to- 
N ratio to soils can lead to short-term N deficiencies as a result of N immobiliza- 
ii. -n 

Residues, however, can be more effective than manures with perennial 
crops. Sanders (1953) reported a 22% yield increase of coffee from manure and 
a 50% yield increase from banana trash mulches for a 10-yr period. Reports by 
Gilbert (1945), Bull (1963), and Robinson and Hosegood (1965) suggest that 
these responses were a direct result of improved soil nutrients, soil-water conser- 
vation, and modulated soil temperatures. Banana trash used as mulch contributed 
more to soil-water conservation and modulated soil temperatures than to supply 
of n 

The combined use of mineral fertilizers and organic inputs increased yields 
and maintained stable yields for the duration of the experiments (Table 3-2). In 
almost all cases, the results of mineral fertilizer plus organic materials were addi- 
tive effects of the two inputs. The application of FYM and CR with mineral N 
and P fertilizers resulted in greatest yields at Kabete, Kenya (Fig. 3-3). 

The application of lime with or without mineral fertilizers tended to 
increase and sustain yield in many cases (Table 3-2). The comparable effects on 
yield for lime and for organic inputs with mineral fertilizers suggests that organ- 
ic inputs may have a beneficial effect in reducing soil acidification. Wong et al. 
(1995) showed that application of FYM and calliandra {Calliandra calothyrsus 
Meissner) primings significantly lowered Al saturation from 80 to 68% on acid 
Oxisols of Burundi. The reduction in exchangeable Al, which was attributed 

Fig. 3-3. Effect of applications ol mineral fertilizer (120 kg N ha- 1 yr- 1 , 52 kg P ha- 1 yr- 1 ), farmyard 
manure (FYM, 10 t ha- 1 yr- 1 ) and returned crop residues (CR) on maize grain yields at Kabete, 
Kenya (S.M. Nandwa, KARI, 1997, unpublished data). Values are 3-yr moving averages. 


mainly to the applied organic materials, increased maize grain yield. The benefi- 
cial effects of organic materials are reviewed in detail by Palm et al. (1997, this 

Results of the long-term experiments emphasize the need for greater use of 
fertilizers (mineral and organic) to remedy the nutrient deficiencies in Africa. For 
intensive and continuous crop production, these inputs should aim at balanced 
application of nutrients and avoidance of soil acidification. 


Our review shows a history of quality research that supports fertilizer use 
in improved crop production. We have learned of the benefits of mineral fertiliz- 
er inputs combined with crop residues, manure, and biological N2 fixation for 
conserving and maximizing nutrient-use efficiency. Yet, less fertilizer is being 
applied per unit land area in Africa than in other regions of the world (Woomer 
& Muchena, 1996; FAO, 1996). 

We technically view the limited adoption of fertilizer use by farmers as a 
symptom of deficiencies in many research trials. The most significant appears to 
be their poor representation of farmers' conditions. Nearly all the research results 
reported in this review were obtained from experiments conducted on research 
stations or at field sites wholly managed by researchers. The studies have 
involved single crops or simple seasonal rotations of a few crops in contrast to the 
intercropping frequently preferred by smallholders. Under smallholder condi- 
tions, it is rarely possible to generate organic inputs to meet the rates reported in 
the research studies. There also is insufficient information on the optimum use of 
small amounts of mineral fertilizers to supplement nutrients contained in organic 
residues that may be imported on the farm. 

Many important practical questions for smallholders thus remain unan- 
swered because they were not built into the designs of these experiments. We pro- 
pose mat in designing experiments to supplement data already generated in past 
research, scientists should trace their research activities to farm productivity. This 
requires involvement of farmers as active partners in all the events of the research 

Only a few studies had inferences to short-term economic optimum rates of 
mineral fertilizer application. Studies seldom consider appropriate fertilizer rates 
to sustain profitability on a long-term basis. Although the average fertilizer con- 
sumption rates in Africa are low, fertilizers are used on cash crops such as cotton, 
coffee, and oil palm that can steadily be marketed on a profitable basis (Vlek, 
1990). If other crops could likewise be effectively marketed, the use of fertilizer 
would increase. This is supported by an example from Nigeria (Vlek, 1993) 
where farmers were willing and able to produce grain for the market and purchase 
the necessary fertilizer when they did not have to compete with subsidized grains. 
If research could demonstrate that using fertilizer on stable food crops was prof- 
itable, the adoption of skills and technologies for better fertilizer use would be an 
integral process of increased agricultural production. 



Adepetu, J.A., and R.B Core) 1976 Organic ph us as a predictor of plant available phospho- 
rus in soils of southern Nigeria. Soil Sci. 122:159-164. 

Acluayi. E.A. 1984. Effects of long-term ammonium sulphate fertilization on the pH and elemental 

distribut ii i linn i i leptl >1 outli tern lis i >il ' mmun. Soil Plant Anal. 

Anderson, CD. 1974. Pota i 

the lOtli Colloquium of tile Int. Potash In 

Inst, Bern Switzerland. 
Arora, Y., and A.S.R. Juo, 1982. Leaching of fertilizer ions in a kaolinitic Ultisol in the high rainfall 

tropics: Leachii in field plots nuclei pint: ai |i il i \ i 

J. 46:1212-1218. 
Bation E id M nomic ai i i ii niemsi ph 

k hi cliff i n 1 il n Mali. Nu lin I I ' I 

Bationo, A., A. Buerkert, M.P. Sedogo, B.C. Christianson, and A.U. Mokwunye, 1995. A critical 

review of crop residue use as soil amendment in the West African semi-arid tropics, p. 
ii In JM Po\ II et a i Livestocl nil ustaii I nutri i i i lin in mix I farm 

n i hi 1 1 ill i i i fri t'ol di I i| In I i 1 'Hi' frica. Addi 

Ababa, Ethiopia. 
Bationo, A., C.B. Christianson, W.E. Baethgen, and A.U. Mokwunye. 1992. A farm-level evaluation 

of nitroiien and P fertilizer use and planting densit\ for near! millet production in Nisei. Fen. 

Res. 31:175-184. 
Bationo. A.. C.B Christianson. and M.C. Klaij. 1993. The effect of crop residue and fertilizer use on 

pearl millet yields in Niger. Fert. Res. 34:251-258. 

AC. Mokwunye. 1991. Alleviating soil fertility constraints to increased crop pro- 

: evaluation of phosphate fertiliz- 
n tropical Africa, p. 283-318. In P.L.G. Vlek and A.U. Mokwunye (ed.) Management of 

nitrogen and P fertilizers m sub-Saharan Africa. Martinus Xijhoff Publ.. Dordrecht, the 

Bosv.'inkle. E. 1991. Residual effects ol phosphorus fertilizers in Kenva. Empire J. Exp. Auric. 

Brodrick, S.J., F. Amijee, J.A. Kipe-Nolt, and K.E. Giller. 1995. Seed analysis as a means of identi- 

fviiiii micronutrient deficiencies of I'IkiscoIiix vul:;cifi\ L in the tropics. Trop. Auric. 

(Trinidad) 72:277-284. 
Bull. R.A. 1963. Effects of mulching and irrigation in some East African coffee soil. A review. 

Turrialba 13:22-27. 
Buresh, R.J., PC. Smithson, and D.T. Heliums. 1997. Building soil phosphorus capital in Africa, p. 

111-149. In R.J. Bureshetal. (ed.) Replenishing soil fertility in Africa. SSSA Spec. Publ. 51. 

SSSA, Madison, WI (this publication). 
Buresh. R.J.. and G. Turn. 1997. Soil improvement by trees in sub-Saharan Africa. Agrofor. Syst. (In 

Cassman. K.G.. P.W. Singleton, and B.A. Linquist. 1993. Input/output analysis of the cumulative soy- 
bean response to phosphorus on an Cltisol. Field Crops Res. 34:23-36. 

Chabalier, P.F., and .1. Pichot. 1979. Utilization of fertilizer nitrogen b\ maize crops in the Ivory Coast. 
p. 3 i-47. /../ Isotopes and radiation in research on soil-plant relationships. Inf. Atomic Energ) 
Agency, Vienna, Austria. 

Christianson, C.B., and P.L.G. Vlek. 1991. Alleviating soil fertility constraints to food production in 
West Africa: Efficiency of nitrogen fertilizers applied to food crops. Fert. Res. 29:21-33. 

Deckers, J. 1993 Soil i 1 i nd environmental problems i iff i I il zones of the devel- 
oping countries in sub-Saharan Africa, p. 37-52. In H. Van Reuler and W.H. Prins (ed.) The 
i [ I mi mi n i i i r in i i i ] Ii n in sul i n Afri renigi m 

Kunstniest Producenten. Leidschendam. the Netherlands. 

:. Organ, of the United Nations, 

Foster. H.L. 1979. Soil fertility in Uganda. Ph.D thesis. ! 'niv. of Newcastle upon Tyne. England. 
Foster, H.L. 1979. Yield responses to lime and potassium fertilizer in different cropping years in 

Uganda. Exp. Agric. 15:193-201. 
Fox, R.L., and B.T. Kang. 1978. Influence of phosphorus fertilizer placement and fertilization rate on 

maize nutrition. Soil Sci. 125:34-40. 


Friend, M.T., and H.F. Birch. 1960. Phosphate responses in relation to soil tests and organic phos- 
phorus. J. Agric. Sci. (Cambridge) 54:341-347. 

I i ii. D.l I'M 1 i i ffi n f sulfur fertilizer appl i r i ( j n West Africa. Fert. 
Res. 29:35-44. 

Friesen, D.K.., A.S.R. Juo, and M.H. Miller. 1982. Residual value of lime ami leaching of calcium in 
akaolinitic Ultisol in the high rainfall tropics. Soil Sci. Soc. Am. J. 46:1184-1189. 

Gamy. F.. G. Guiratkl. ami Y. Dommergues. 1978. Effect of straw incorporation on the yield and 
nitrogen halm - in il ncl oil-j n mill i croppin i m eneaal Plant Soil 


Gerner, H., and A.U. Mokwunye (ed.). 1995. Use of phosphate rock foi sustainable agriculture in 
West Africa. Misc. Fert. Stud. 11. Int. Fert. Dev. Ctr. for Africa, Lome, Togo. 

Gigou. .1.. and I i hi ii I > ' ti \ i i i 1 riii i i n in i In i i in i i 

ern Cameroon, p. 49-65. In Isotopes and radiation in research on soil-plant relationships. Int. 
Atomic Energy Agency. Vienna. Austria. 

Gilbert, S.M. 1945. The mulching of coffee Arabica. East Afr. Agric. J. 11:75-79. 

Giller, K.E., G. Cadisch, C. Ehaliotis, E. Adams, WD. Sakala, and P.L. Mafongoya. 1997. Building 
soil nitrogen capital in Africa, p. 151-192. In R.J. Buresh et al. (ed.) Replenishing soil fertil- 
ity in Africa. SSSA Spec. Publ. 51. SSSA, Madison, WI (this publication). 

GHIer. K.E.. J.F. McDonagh. ami G. Cadish. 1994 Can biological nitrogen fixation sustain agricul- 
ture in the tropics? p. 173-191. Ini.J.K.. Syers and D.L. Rim men eel. i Soil science and sustain- 
able land management in the tropics. CAB Int., Wallingford, England. 

IAEA. 1975. Root activity patterns of some tree crops. Tech. Rep. Ser. 170. Int. Atomic Energy 
Agency, Vienna, Austria. 

ICRAF. 1997. Annual report 1996. Int. Ctr. for Res. in Agrofor., Nairobi, Kenya. 

ICRISAT. 1989. Annual report 1988. Int. Crops Res. Inst, for the Semi-Arid Tropics, Sahelian Ctr., 
Niamey, Niger. 

JRAT. 1974. Rapport annuel 1972 de l'JRAT. LAgronomie Tropicale 29:464-467. 

I i' E.O I "' ii i 196-1 >il iiln ml nun ent si i in di mi ent soil- vegetation 

systems in a tropical rain-forest environment. Trop Ague. (Trinidad) 41:1 J!')- 1 39. 
na, B., R.A. Swinkels, and R.J. Buresh. 1997. Agronomic and economic evaluation of organic and 
inorganic sources of phosphorus for maize. Agron. J. 89:597-604. 
n i ' I Ini i nutrient nun nent: The u f organi i I mi i il I rtili i 

89-105. In H. Van Reuler and W.H. Prim; (ed.) The role of plant nutrients for sustainable crop 
production in sub-Saharan Africa. Vereniging van Kunstmest Producenten. Leidschendam. the 
les, P.A., J.B.D. Robinson, and J. Wallis. 1960. Fertilizer, manure and mulch in Kenya coffee grow- 
ing. Empire J. Exp. Agric. 28:335-352. 
>. .-vS.R. 1981. Chemical characteristics, p. 51-79. In D.J. Greenland (ed > Characterization of soils 
in relation to their classification and management for crop production: Examples from some 
areas of the humid tropics. Oxford Unp Pi Oxford England. 
>, A.S.R., A. Dabiri, and K. Franzluebbers. 1995. Acidification of a kaolinitic Alfisol under con- 
tinuous crop] i ith nitrogen fertili itioi i\ >i Plant Soil 17:245-253. 

Kang, B.T., and O.A. Osiname. 1985. Micronutrient problems in tropical Africa, p. 131-150. In 
P.L.G. Vlek (ed.) Micronutrienl in tropi I food crop pi luction laitinu ijhoff/Di ' 
Junk Publ., Dordrecht, the Netherlands. 

Kenya Agricultural i ! i In titutt 1"" 1 rtili i i ( u i t ndations Vol. 1-22. Fertilizer 

Use Recommendation Pro|ect. Kenya Agric. Res. Inst, Nairobi, Kenya. 

Laryea, K.B., M M n i mil I id k I''' i m I m ,| r m m n \lfi 1 ind \ il 1 ii 
the semiarid tropics, p. 267-292. In R. Lai and B.A. Slew air (ed.i Soil management: 
I inn ill ill i i 111 ill nd environmental qualit 'Ml' 1 i Raton. FL. 

Le Mare, P.H. 1959. Soil fertility studies in three areas of Tanganyika. Empire J. Exp. Agric. 
I H. 1974. 1 i i i i i i i i 1 Ii i n ill hat md relation 

i t\ i I i, I in il nstituents in Tanzania. J. Agric. Sci. (Cambridge) 83:47-56. 

Le Mare, P.H. 1981. Phosphorus sorption and release, p. 97-134. In D.J. Greenland (ed.) 

Characterization of soils in relation to their classification and management for crop produc- 

i 1,1 i 1 i i l 1 I II mi II < 1 I 1 I ' ' I I Li I a I 

1....' Mare. P.H. 1984. Limitations imposed b\ nutrient supph in tropical African soils, p. 357-361. fa 
D.L. Hawksworth (ed.) Advancing agricultural production in Africa: Proc. of CAB's First 
Scientific Conf. Arusha. Tanzania. 12-18 Feb. I '184. C onimonwealtli Agric. Bui., England. 


Lungu, O.I., J. Temba, B. Chirwa, and C. Lungu. 1993. Effects of lime unci farmyard manure on soil 
acidity and maize growth on an acid Alfisol from Zambia. Trop. Agric. (Trinidad) 

MacColl, D. 1989. Studies on maize (Zea mays L.) at Bunda, Malawi: II. Yield in short rotation with 
legumes. Exp. Agric. 25:367-374. 

McWalter, A.R and P II ■> im I 1 ( \ long-i mi i I iti n; I md id nuri trial in Uganda. Exp. 
Agric. 12:305-317. 

Mokwunye, A.U. 19 < Mi ill i ferti rs in Nigt i nn il 11 I i tuition of three phos- 

ui ppliecl i i i i I i ( i I ii I i 

Mokwunye, A IT le . lager, and E.M In I ' Tori i nd m n nin i 

i ii i West African II istainable d lopmen n I it Stud. 14. Int. 

Fert. Dev. Ctr.-Africa, Lome, Togo. 

Mughogho, S.K., A. Bationo, C.B. Christianson, and P.L.G. Vlek. 1986. Management of nitrogen fer- 
tilizers for tropical African soils, p. 117-172 In U. Mokwunye and RL.G. Vlek (ed.) 
Management of nitrogen and phosphorus fertilizers in sub-Saharan Africa. Martinus Xijhoft 
Publishers, Dordrecht, the Netherlands, 

Mughogho, S.K., C.B. Christianson, J.M. Stumpe, and P.L.G. Vlek. 1990. Nitrogen efficiency at three 
sites in Nigeria as affected by N source and management. Trop. Agric. (Trinidad) 67:127-132. 

Mukurumbira, L.V1. 1985. Effecl if ral iff rtili nil n m pi iou gi un legume crop on 
maize yields. Zimbabwe Agric. J. 82:177-179. 

Nagy, J.G., H.W. Ohm, and S. Sawadogo. 1990. Farmer-researcher collaboration in developing crop- 
in i i ii i it in lve I ii i lei i in i i i ii Ii hi i 1 n rid md 1 rtili i t h- 
I i rghinn in Burkina 1 * 1 1 161- 169. 

Nyakatawa, E.Z. 1996. Rain water and soil fertility management for sustainable cropping on sandy 
soils in semi-arid South-East Lowvekl of Zimbabwe. J Sustain. Agric. 7:19-34. 

Osiname. O.A. 1979 lai i n t I lion iliz ion in differem I I i t 

ern Nigeria. Niger. .1. Agi ic. Sci. 1 :9-l I . 

in II 1 ii i I 1 n I 1997. Combined ii f organ nd inorganic nutrient 
sources for soil fertility maintenance and replenishment, p. 193-217. In R.J. Buresh et al. (ed.) 
i | tun hi i rtilitj in Africa. SSSA Spec. Publ. 51. SSSA, Madison, WI (this publica- 

Pi ii ' 197 i iuiiiu ' cereal d .1 in - h n Re] ibliqut i vl Ii render essai de syn- 
thase. L'Agronomie Tropicale 28:751-766. 
i I i in i i I ii frica. p. 41 -6 1 . /// I \1 magement of acid 

tropical soils for sustainable agriculture. IBSRA.Y1 Inaugural Workshop. Yurimaguas. Pern. 

and Brasilia Brazil. 24 Apr.- ? May 1985. Int. Board for Soil Res. and Manage., Bangkok, 

Fieri, (.". 1995. Long-term soil management experiments in semiarid Francophone Africa, p. 225-266. 

In R. Lai and f ' i rt i i il inana ment: Experimental I i i i i in ilitj nd 

environmental quality. CRC Press, Boca Raton, FL. 
Piha. M.I. 1993. Optimizing fertilizer use and practical rainfall capture in a semi-arid environment 

with variable rainfall. Exp. Agric. 29:405-415. 
Rajan, S.S.S., J.H. Watkinson, and A.G Sinclair. 1996. Phosphate rocks for direct application to soils. 

Adv. Agron. 57:77-159. 
Robinson, J. B D md P.H. H 19 Effect rganic mulch on fertilii of a latosolic coffee 

soil in Kenya. Exp. Agric. 1:67-80. 
Sanchez, P.A. 1976. Properties and management of soils in the tropics. John Wiley & Sons, New 

Sanchez, P.A., K.D. Shepherd, M.J. Soule. F.M. Place. R.J Buresh, A. -M.N. Izac, A.U. Mokwunye, 
F.R. Kwesiga. C.G. Xdiritu, and PL. Woomer. 1997b. Soil fertility replenishment in Africa: 
An investment in natural resource capital p. 1-46. hi R..I. Buresh et al. (ed.) Replenishing soil 
fertility in Africa. SSSA Spec. Publ. 51. SSSA, Madison, WI (this publication). 

Sanders, F.R. 1953. A review of the fertilizer and mulch experiments on Coffcti arabica in 
Tanganyika. East Air. Agric. J. 19:44-45. 
i vl I9( I i fertili laments in Western Tanzani I. Ami Sci (Cambridge) 

in MA. 1971. T long-tei ff'ects of fertilize! farmyard main md I at Mwanhala, west- 
ern Tanzania. East Afr. Agric. For. J. 3 7:8- 1 4 


> H ( llltUl Illfkl 111 il nit I 

male de Cote d'lvoire, Abidjan, Cote d'lvoire. 
Sedogo, P.M., B.V. Bado, V. Hien, and F. Lompo. 1991. Utilisation efficace des engrais azotes pour 

i i i i i i I In i\ I in 11 i i i ii i I I I 

A.L". Mokwiinye led. I Alleviating soil feitilitv constraints to increased crop production in 
West Africa. KJuwer Academic Publ., Dordrecht, the Netherlands. 
Singh, B.R., and H.C. Goma. 1995. Lou in il t'ertilit in i n i nments in Eastern 

Africa, p. 347-382. In R. Lai and B.A. Stewart led.) Soil management: Experimental basis for 
ustainabiiil nd envi nmental qualil < 1 ' I l 1 , Raton. FL 

ii L.. and 1 I i n n I Eft i i ulizei use on a ferruginous soil 

(Haplustalf) in Nigeria. Exp. Agric. 15:257-265. 
Smaling, E.M.A., S.M. Nandwa, and B.H. Janssen. 1997. Soil fertility in Africa is at stake, p. 47-61. 
In R.J. Buresh et al. (ed.) Replenishing soii fertility in Africa. SSSA Spec. Publ. 51. SSSA, 
Madison, WI (this publication). 
bulo II > .ii O.A. Osin in I >f," oil prop ti ind ropyi Id under , nitinuous cultivation 
mi hit in ii a i,ii in i ' / I L i i il I ii i I in i I i 

opment in the tropics. A.A. Balkema. Roterdam. the Netherlands. 
Ii. H. 1 1» i mi ii 1 i i ii n I I i ii ulii n f the cooking type) 

during the \ uptake. M.S. the- 

sis. Makerere Univ., Kampala, Uganda. 
Ssaii. [-1. I 1 !-. 1 !!. Initial and residual effects of nitrogen fertilizers on gram yield of a maize/bean inter- 
i i ,i ii i I i mi i I i i i i i i n i ill i i plied nit ii i i ' 

Ill ii i i > i ! I n ii pli lion ii I i ii i i i ill i level on nodulation, 

i \uh ind liniti igen fi i mi ol ilu 1 ni iih frop :ri 1 1 inid d} ( l> 109 

iiii ' > in FN i i fallow ti on ni i i i i n i id 1 i i t i ii- mda clay loam in 

South Uganda. J. Agric. Sci. (Cambridge) 68:391-403. 
tephens. D. I Ii n I nanure and trace elements i ntinuous cropping rota- 

tions in southern and western Uganda. East Afr. Agric. For. J. 34:401-417. 
Stoorvogel, J.J., and E.M.A. Smaling. 1990. Asses in n il oil mini nt I pi tion in sub-Saharan 
Africa. 1983-2000. Rep 28. Winand Staring Or. for Integrated Land, Soil and Water Res., 
v mil n iht 1 i 1 no 
i i I I l 1 I.A Sn lin id B.H. Jan ml > 1 tin i I nutri nt balances in Africa 

I dill i n I 1 i| I il I I il Fert. 1 

Swift, M.J., P.D. Seward, P.G.H. Frost, J.N. Qureshi, and F.N. Muchena. 1994. Long-term experi- 
ments in A.f i i i I i i in ill i i ni inible land use under glob il i n p. 

I r I I til i i I i 

ecological si i CA1 lia \ i lin ;fc i i England. 

Tag, P., A.J. Okalo, and K. Zschernitz. 1972. The FAO fertilizer programme in Kenya. In Proc. 12th 
in ting i I i in ii i ii i niinitl i il F iiili i (i < i | utiition, Morogoro, 
Tanzania. 16-18 May 1972. 
Traore S.. and P..I. Harris. 1995. Lonti-term fertilizer and crop residue effects on soil and crop yields 
in the savanna region of Cote D'lvoire. p. 141-180. In R. Lai and B.A. Stewart (ed.) Soil man- 
in nl | i mi in il I i t i I 1 ii il il i n i ii ii n i ni I i i i i CI ' Boca 

Raton, FL. 
Tukahirwa, E.M. 1992. Uganda: Environmental and natural res 
Issues and options: II. Documentation. Makerere Univ. 
Kampala. Uganda. 
Uyovbisere, E.O.. and G Lombiii 1991 Efficient I rtiliz in foi increa I op production: The 

sub-humid Nig n so, ri nee Fert. Res. 29:81-94. 
Van Reuler, H., and W.H. Prins. 1993. Synthesis, p. 3-1 1. fa H. Van Reuler and W.H. Prins (ed.) The 
i > I mi in h i in i i ii ii i i Mi du i i in in harm * m ni in 

van Kunstniest Producenten. Leidschendani. the Netherlands. 
Vlek. P.L.G. I')"!!. The role of fertilizers in sustaining agriculture in sub-Saharan Africa. Fert. Res. 

Vlek, P.L.G. 1993. Strategies for sustaining agriculture in sub-Saharan Africa, p. 265-277. In J. 
.' ml I', I ,n,l I' Lai (ed.) 1 hnolosiit i „ u i in bl i mi ulturt m tin tropi s 
Publ. 56. ASA, CSSA, and SSSA, Madison, WI. 
Wapakala, W.W. 1976 Change,, in some chemical properties of a red clay soil: Results oflong-term 

fertilizer and rotational trials in central Kenya. East Afr. Agric. For. J. 42:201-218. 
Warren, G. 1992. Fertilizer phosphorus sorption and residual value in tropical African soils. NRI Bull. 
i7. Nat. Resour Inst.. Chatham. England. 


Wong, M.T.F., E. Akyeampong, S. Nortcliff, M.R. Rao, and R.S. Swift. 1995. Initial responses of 

ii ii 1 i i ii i in i ni ii ii 1 1 ii i 1 in m in itli [ li 

tion of manure or tree prunings to an Oxisol in Burundi. Plant Soil 171:275-282. 
' ii ( L I 1 ' ' ' hi in I i i n| 19 i ii 1 i i ii ii i i in the East African 

Highland h biological fixati I 

(in press). 
i I L i i I I i / i Iri i 

duction in tropical Africa. Afr. Crop Sci. J. 4:503-518. 
Yamoah, C.F., J.R. Burleigh, and V.J. Eylands. 1992. Correction oi acid infertility in Rwandan 

Oxisols with lime from an indigenous source for sustainable cropping. Exp. 

Yost, R.S., E.J. Kamprath, E. Lobato, and G. Naderman. 1979. Phosphorus response of corn on an 

Oxisol as influenced by rates and placement. Soil Sci. Soc. Am. J. 43:338-343. 

A Fertilizer-Based Green 
Revolution for Africa 

Marco A. Quifiones, Norman 
and Christopher R. Dowsweli 

Sasakawa Africa Association 
Mexico D.F., Mexico 


On average, <5 kg ha- 1 of mineral fertilizer nutrients are applied to food crops in 
sub-Saharan Africa, the lowest rate in the world. Population pressures have caused tradi- 
tional systems of shifting cultivation to break down. Soil nutrients are being depleted at an 
alarming rate, leading to environmental degradation and food insecurity. To reduce pover- 
ty and assure food e urit; ifri ,n igi ii ulture must grow at 4 to 5% per year, more than 
twice the rate of recent decades. This growth is unattainable without using significantly 
greater quantities of mineral fertilizers. Excellent agronomic response to fertilizers has 
been observed in >600 000 half-hectare, on-farm demonstration plots in the major food 
crops in 12 countries. In Ethiopia, with the increased use of mineral fertilizer, improved- 
seed, better extension advice, and favorable rainfall over most of the country; record har- 
vests of the major food crops have been achieved over the 1 995- 1 996 and 1 996-1997 sea- 
sons. The country has become food self-sufficient and advanced closer to the goal of food 
security. Attention must be given to improving the efficiency of use and supply of fertil- 
izers to make them more affordable to farmers. Equally important is the role of agricul- 
tural research and development in devising technologies and strategies that ensure the sus- 
tainability of food production to meet the demand of an ever-increasing population. 
Priority on the agenda should be given to developing nutrient management practices in 
which all nutrient sources are judiciously integrated and losses tc 

Between 1997 and 2020, population in su I thai n ri i (hereafi i ret rre I 
as Africa) will more than double to over 1.1 billion people (Dyson, 1995; Rose- 
grant et al., 1995). Investments by national governments and the international 
community have been insufficient to arrest poverty, assure food security, and 
reduce environmental degradation in this continent. Indeed, if present trends con- 
tinue, food insecurity, malnutrition, and resource degradation will increase, and 
by 2020, it is conceivable that Africa will need to import between 50 and 70 mil- 
lion tonnes per year of foodstuffs (mainly cereal grains) to meet the demands of 

Copyright © 1997 American Society of Agronomy and Soil Science Society of America, 677 S. 
Segoe Rd, Madison, WI 53711, USA. Replenishing Soil Fertility in Africa. SSSA Special Publication 


the increased population (Dyson, 1995; GCA, 1996). Almost certainly, Africa 
will not have the economic resources to procure such huge volumes of food on a 
commercial basis nor will the international community be willing to provide it as 
concessional sales or food aid. 

Since most of the poor in Africa are rural, and agriculture is their mainstay, 
it follows that agricultural development must be the central strategy for econom- 
ic growth and poverty alleviation (Birdsall, 1995). It also is important to stress 
that hunger, poverty, and environmental degradation in Africa are intimately cor- 
related, and that any action to reduce poverty and hunger will assist in minimiz- 
ing environmental degradation as well (Cleaver & Schreiber, 1994). 

The critical role of agriculture in Africa's development is not universally 
accepted by many governments. Consequently, in most countries, agriculture has 
been growing more slowly than population, which has resulted in decreasing in- 
comes in real terms. The major cause of the poor performance of the agriculture 
sector is not a dearth of improved agricultural technology that can empower the 
small-scale farmer to increase productivity. Rather it is poor economic and agri- 
cultural policy; inadequate investment in infrastructure and rural education; in- 
sufficient agricultural services, such as research, extension, credit, input supply, 
and marketing; and low investments in rural health care (Cleaver & Donovan, 

In most regions of the world, increases in food supply over the past two 
decades have resulted mainly from raising yields. The only major exceptions are 
Africa and the Cerrado of Brazil, where most of the growth in production has 
occurred because of expansion of the cultivated area (Borlaug & Dowswell 1994; 
Dyson, 1995). It is widely perceived that technology-based agriculture has large- 
ly bypassed Africa. Where land is plentiful, slash and burn shifting cultivation 
persists, and this is still common in much of Africa. However, where population 
pressures have reduced the fallow period, a sedentary low-yield agriculture has 
arisen. But no matter what the variations in the agricultural system, the common 
base is that plant nutrients are the minimum factor for crop production (Jansson, 
1995). Traditional agriculture results in mining soils of plant nutrients by remov- 
ing crop residues, leaching, and soil erosion (Smaling et al., 1997, this publica- 
tion). According to Stoorvogel and Smaling (1990), about 200 million ha of crop- 
land in Africa have lost 660 kg N ha" 1 , 75 kg P ha" 1 , and 450 kg K ha- 1 during 
the last 30 yr, primarily by removing crop harvests (Bumb, 1995; Sanchez et 
al.,1997, this publication). These figures amount to a loss of plant nutrients in the 
range of about 8 million tonnes N-P-K annually. 

Traditional fanning systems in Africa are responsible for the loss of 4 mil- 
lion hectares of forest that are cleared annually to give room or substitute for the 
cropland that has become unproductive because of nutrient depletion. This prac- 
tice is leading to disastrous environmental consequences, such as soil erosion, 
weed invasions, impoverished postfire-climax vegetative ecosystems, and loss of 
biodiversity (Borlaug & Dowswell, 1995). The solution is clearly not to expand 
food production horizontally to keep pace with population growth at the cost of 
environmental degradation. Instead, the solution is to provide adequate soil nutri- 
ents by increasing the use of mineral fertilizer, combined with organic inputs that 
build up organic matter in the soil, and the complementary practices of using 


improved seeds and proper plant population, weed control, and other cultural 

Fertilizers have played and will continue to play an important role in 
increasing the food supply for future generations. It is estimated that around 50% 
of the annual global food harvest comes from the application of mineral N fertil- 
izer alone (Dyson, 1995). Thejudicious use of mineral fertilizers can play a crit- 
ical role in preventing resource degradation that results from nutrient mining, and 
from the exploitation of fragile lands or the clearing of habitat-rich forests. In 
Africa, fertilizer consumption on food crops is the lowest in the world — probably 
no >5 kg ha" 1 of nutrients, when fertilizer use on cash crops is subtracted from 
aggregate statistics. 

Increased fertilizer use in Africa can create a win-win situation, by pro- 
moting more efficient crop production and reducing soil degradation. Mineral 
fertilizers should be at the core of strategies to restore soil fertility and raise crop 
productivity, although their use should be a part of integrated systems of nutrient 
management in which organic fertilizer sources are included. Organic sources of 
nutrients, however, will be complementary to the use of mineral fertilizers, and 
not the other way around. Exclusive use of organic fertilizers will increase food 
production at best by 2% yr 1 (Hiyami & Ruttan, 1985), well below the popula- 
tion growth rate, and not even close to the 5 to 6% required to reduce poverty and 
assure food security. The World Bank's 1989 long-term perspective study (1990 
to 2020) to permit gradually improved food security and increased rural incomes 
set agriculture growth in Africa at 4% a year (Cleaver & Schreiber, 1994). 
Because the overall performance of Africa for 1990 to 1996 did not attain the 4% 
target, many people including Cleaver (1996, personal communication) say that 
5% per year agricultural growth is probably necessary to have a significant 
impact on poverty reduction. With population growth around 3% per year, this 
would be only a 2% net increase in per capita production. 

It also is important to mention that sources of organic manure are limited 
in most African countries. Even in Ethiopia, where livestock numbers are signif- 
, manure is primarily used as a cooking fuel and rarely to improve the fertil- 
ity of the soil. Moreover, use of other organic sources, such as green manures, 
presupposes growing the manure crop at the expense of a food or cash crop 
(Giller et al, 1997, this publication). Finally, alley cropping and agro forestry 
approaches to maintaining soil fertility are knowledge-intensive, nutrient man- 
agement systems that have met with limited success, especially where poverty 
and hunger force farmers to employ desperate short-term survival strategies that 
take precedence over longer term sustainability practices. Hence, efforts should 
be made to increase the efficient use of mineral fertilizers through sound policies 
and education, to attain economic growth and food security targets while mini- 
mizing the damage to the resource base. 


The term Green Revolution has been much misunderstood since it was first 
coined by William Daud, former administrator of US AID, some 30 yr ago. We 


define the Green Revolution as the beginning of a new era for agricultural 
research and development in the third world, one in which modern principles of 
genetics and plant breeding, agronomy, plant pathology, entomology, cereal tech- 
nology, and economics have been applied to develop higher yielding technologies 
appropriate to the conditions of local farmers (Borlaug, 1988). The Green 
Revolution concept to produce more food by increasing the productivity of the 
more favorable agricultural lands becomes especially relevant as the per capita 
availability of arable land declines. It is to these principles that we adhere when 
referring to a Green Revolution for Africa. 

Sasakawa-Global 2000 (SG 2000) began its agricultural projects in Africa 
in 1986. Our mission was to contribute towards the attainment of food security 
through the adoption of productivity-enhancing technologies by small-scale 
farmers. The SG 2000 projects have been funded since their inception by the 
Sasakawa Foundation, recently renamed as the Nippon Foundation, and they are 
enthusiastically supported by former U.S. President Jimmy Carter. Projects are 
currently in operation in Ghana, Benin, Togo, Nigeria, Guinea, Mali, and Burkina 
Faso in West Africa and Ethiopia, Eritrea, Tanzania, Mozambique, and Uganda 
in East and southern Africa. Similar projects also were operated previously in 
Sudan and Zambia. 

At the outset, it is important to make clear that SG 2000 conducts the 
majority of its program activities with — and through — national research and 
extension organizations. We do not operate separate parallel programs, and SG 
2000 only has six field directors to supervise operations in 12 countries. The core 
of the SG 2000 projects are dynamic field testing and demonstration programs for 
the major food crops in which improved technology exists but for various reasons 
was not being adequately extended to farmers (Borlaug & Dowswell, 1995). The 
SG 2000 projects work under the leadership of the national extension depart- 
ments of the relevant ministries of agriculture. Practically all the technical exten- 
sion staff from those departments are thoroughly involved in the planning, imple- 
mentation, and monitoring of SG 2000 field programs. At the regional or state 
level, this leadership role is transferred to the regional or state offices of agri- 
culture, who appoint district or zonal coordinators for the field program. The dis- 
trict offices of agriculture assign the frontline extension staff who assist the par- 
ticipating farmers in the establishment of their demonstration plots and help in 
other project-related field activities (such as field days and postharvest work). 

The selection of participating farmers is done by the frontline extension 
staff in collaboration with their supervisors. The participation by farmers in the 
establishment of SG 2000-sponsored demonstrations is voluntary. Farmers dis- 
cuss and agree on the conditions of participation, which usually involve agree- 
ment on the part of the farmers to follow application of the recommended kinds 
and amount of production inputs followed by proper cultural practices. Special 
efforts are made to engage women farmers, particularly in countries where food 
production undertakings are the responsibility of women. 

SG 2000 works mainly in agroecologies known for their high agriculture 
potential, and its field program emphasizes the application of research-led infor- 
mation that can bring about dramatic increase in productivity. Working with 
national extension services during the past 1 1 yr, small-scale farmers have grown 


>600 000 demonstration plots (0.25-0.5 ha). Most of these plots (known by dif- 
ferent acronyms depending on the country) have been concerned with demon- 
strating improved technologies in maize (Zea mays L.), wheat {Triticum aestivum 
L.; T. turgidum L. var. durum), sorghum [Sorghum bicolor(L.) Moench], cassa- 
va (Manihot esculenta Crantz), grain legumes, barley (Hordeum vulgare L.), 
potato (Solarium tuberosum L.), and, in the case of Ethiopia, tef (Eragrostis tef 
Zucc). Approximately two-thirds of these plots have been planted to maize, 
either as a monocrop, or in various intercropping patterns with cassava or grain 
legumes, or in several multiple cropping patterns with velvet bean (Mucuna 
pruriens var. utilis) and other green manure crops. 

The improved technological packages taken to farmers by frontline exten- 
sion staff, with support from SG 2000, are derived from national and internation- 
al research systems and are upgraded as new research information or cultivars 
become available. The packages of improved crop management practices being 
recommended include (i) the use of the best available commercial cultivars or 
hybrids, (ii) improved agronomic practices that assure proper rates, dates and 
methods of planting, timely weed control, efficient use of available soil water, 
and when needed, crop protection chemicals, (hi) proper application at moderate 
levels of appropriate fertilizers to restore plant nutrients in the soil, and (iv) 
improvement of on-farm storage structures and methods for harvested grain, both 
to reduce postharvest losses and to extend the marketing season by safely hold- 
ing stocks until prices are more favorable. 

One distinctive feature of the SG 2000 technology-transfer approach is the 
size of the demonstration plot, which is usually between 0.25 and 0.5 ha, and con- 
stitutes the actual site on which improved farming practices are appraised by the 
farming community. To add to the economic realism, participating farmers are 
asked to pay the commercial cost of inputs used to conduct the demonstrations on 
their land. While there are differences among SG 2000 project countries, the 
trend is for farmers to pay 50 to 100% of the cost of inputs prior to planting, to 
help ensure that they become stakeholders from the beginning, with a sense of 
ownership and obligation for repayment, instead of fostering dependence on ex- 
ternal aid. After one or two seasons, a participating farmer is graduated and ex- 
tension workers move on to new farmers who are enrolled in the demonstration 

The larger size demonstration plot recommended by SG 2000 not only 
allows farmers to make a more realistic appraisal of the recommended technolo- 
gy but also affords them with a clear measure of the economic returns on their 
labor and capital. In contrast, in the pervasive World Bank-supported Training 
and Visit extension system the demonstration plot is between 0.005 and 0.01 ha. 
With the larger test plots, farmers measure increased yields in terms of 100 kg, 
not kilograms. The SG 2000 demonstration plot approach has gained wide recog- 
nition and approval by extensionists as a very effective tool in the diffusion of 
improved technologies to farmers. 

Simultaneously with the dynamic field demonstration program, SG 2000 
project staff strive to get the attention of policy makers, since we believe that 
international assistance should be used only as a catalyst, not as a substitute for 
national action. The SG 2000 philosophy is that once the technology has been 

Table 4- 1 

. Average mai: 

ze yield in demonstration and farmers' pk 


ize yie 



t plots 


1987 1988 

1989 1990 







2.8 4.0 

3.3 3.2 




5.2 5.2 






2.8 2.4 







t Source, 

Sasakawa-Global 2000 files 


plots farmers'plots 

convincingly demonstrated to African fanners and governmental leaders, they 
should pick up the lessons. In other words, Africans should make the decisions as 
to their development strategies, although the donor community can and should 
assist African countries in developing national capacity. 

The SG 2000-supported demonstration plots are strategically located and 
are reinforced by well-managed field days and information campaigns to gener- 
ate wider scale awareness and interest, not only among neighboring farmers but 
also among key officials at different levels of government. It is not surprising to 
have ministers and even heads of state attending field days and engaging in con- 
structive dialog with farmers. Parallel to the field demonstrations, SG 2000 staff 
work closely with national government leaders in agriculture-sector planning and 
policy formulation to enable farmers to continue using the improved technology. 

In virtually all of the SG 2000 project countries during the past 1 1 yr, 
demonstration plot yields have been two to three times higher than those obtained 
in the control plots or in traditional farmers' fields. Table 4 — 1 shows the yield per- 
formance in maize obtained by participating farmers in selected SG 2000 coun- 
tries. Thousands of field days have been organized and attended by hundreds of 
thousands of farmers who are eager to learn about the success of their fellow 
farmers. A great deal of copying from farmer to farmer also is commonly 
observed. This demonstrates that the technology is not only easily understood and 
implemented but also profitable. Another feature we have observed is that when 
a farmer innovates in one crop and understands the production principles, similar 
innovations in other crops will soon follow. 

If we were to single out one country that vividly demonstrates the potential 
for a Green Revolution in Africa, Ethiopia would be the example for others to fol- 
low. Like most countries in Africa, Ethiopia is predominantly rural and heavily 
dependent on agriculture. Agricultural production accounts for 55% of the gross 
domestic product, and 80% of the population make their livelihood from agricul- 
ture. This sector is dominated almost entirely by small-scale, resource-poor farm- 
ers who produce 90 to 95% of all cereal grains, pulses, and oilseeds, and 98% of 
the coffee {Coffea arabica L.; Central Statistical Authority, 1996b). Cereals ac- 
count for nearly 84% of the total cultivated area and nearly 70% of the caloric 
intake of the Ethiopian population (Central Statistical Authority, 1996a). The 
most important cereal grain crops are tef, maize, sorghum, barley, and wheat in 
that order. 


Table 4-2. Fertilizer imports in Ethiopia.' 

Year DAP Urea N 

Source. National Fertili; 

As in other African countries, Ethiopian agriculture stagnated during the 
1970s and 1980s, and the country was facing chronic difficulties to feed its ever- 
increasing population, which presently is estimated at around 55 million people. 
Annual food imports were close to 1 million tonnes during the last 10 yr, mostly 
as food aid (mainly wheat). This bleak food-aid dependency has experienced a 
remarkable reversal during the past 2 yr, partially because of the support provid- 
ed by the Government of Ethiopia led by Prime Minister Meles Zenawi to put 
agriculture at the forefront of economic development. 

The Ethiopian experience has resulted from a combination of a political 
leader's vision and a fortuitous encounter with the field activities of SG 2000, 
operating in Ethiopia since 1993. During a visit of former U.S. President Jimmy 
Carter in September 1994, an invitation was made to Prime Minister Meles (then 
President of the Transitional Government) to accompany Mr. Carter on a field 
inspection tour of some project sites. Mr. Meles was clearly impressed by the 
demonstration plots grown by the farmers, in which improved seed and mineral 
fertilizers were used in combination with other improved husbandry practices. 
The plots promised to yield three to four times the average yields obtained in the 
area. That watershed visit was the beginning of what is now known in Ethiopia 
as the Intensified Extension Campaign, which is fully backed and supported by 
the government at all levels. This campaign entails the establishment of 0.5-ha 
demonstration plots on farmers' fields in all food crops for which technology is 
available from the national research system and implemented on a massive scale 
covering all agricultural districts of the country. Hundreds of field days are care- 
fully organized to spread the message among the farming community about the 
benefits of adopting improved technology. 

As the result of this campaign, fertilizer imports into the country have 
increased from 47 000 tonnes N and P in 1993 to 137 000 tonnes N and P in 1996 
(Table 4-2). Farmers have been using diammonium phosphate (DAP) fertilizer 
almost entirely for >20 yr, while urea as a source of N is only a recent introduc- 
tion and still needs to be popularized. Consumption of mainly DAP creates an 
imbalance of the N-to-P nutrient ratio. Although most agricultural soils in 
Ethiopia are deficient in both N and P, and present fertilizer research recommen- 
dations across the country range from 70 to 100 kg N ha" 1 and 20 to 30 kg P ha" 1 , 
depending on soil type and specific crop, farmers still depend almost exclusively 
on DAP application. This then leads to an imbalance of insufficient application 
ofN per unit of P (Table 4-2). 


-1996 area and production for major crops of private 

ea Total production 

1994-1995 1995-1996 Change 1994-1995 1995-1996 Change 

Tigray 506 481 

Afar 17 24 

Amhara 2608 2933 

Oromiya 3138 3625 

Soraalie NS 60 

Benishangul-Gummez 60 96 

SNNPR 599 698 

Gambela 6 10 

Harari 5 4 

Addis Ababa 10 10 

Dire Dawa 10 7 

Total 6960 7949 

t Source, Central tati tical \utliorin (1996a) 

J NS, not stated. 

j Southern Nations, Nationalities People's Regioi 

With the increased use of mineral fertilizer during the 1995-1996 and 
1996-1997 crop seasons, Ethiopia recorded its highest harvest of the major crops 
ever in history. This came about not only as a direct result of increased use of 
mineral fertilizers but also increased use of improved seed, better extension 
advice, and favorable rainfall over most of the country (Table 4 — 3). As can be 
observed, during the 1995-1996 season, there was an increase of 14% in total 
area, 32% in production, and 15% in average yields over 1994 — 1995. Preliminary 
production estimates of the 1996-1997 season indicate a second consecutive pro- 
duction record for Ethiopia. The 1996-1997 production estimate represents a 
21% increase over 1995-1996 revised estimate of 9.7 million tonnes. According 
to a FAO/WFP (World Food Program) Crop and Supply Assessment Mission 
who conducted a preharvest crop assessment for the main season; a combination 
of factors contributed to the bounty of the harvest, including timely and consis- 
tent good rains (50% of the production increase because of this), increased area 
under production (6% over last year) and 20% of the production increase because 
of the government extension program (FEWS and EU-LFSU, 1996). 

All of the cereal crops except wheat have experienced substantial produc- 
tivity gains (Table 4- A ). Wheat yields were reduced because of a stem rust epi- 
demic that completely destroyed the most widely grown wheat cultivar, Enkoy. 
The experience with stem rust points out the need to strengthen the national 
wheat research and variety release system in order to (i) broaden the genetic base 
for disease resistance in future cultivars, and (ii) implement a functional disease 
surveillance system that will identify changes in pathogen populations early 
enough so that new resistant cultivars can be released and multiplied to substitute 
for those that have become susceptible. 

The yield gains for millet [Pennisetum glaucum (L.) R. Br.] and oat (Avena 
sativa L.) were large (Table 4 — 4). They, however, are relatively minor crops that 

■s of 1994-1995 aiiii 1995-1996 area, production, and yield of cereal crops for private peasant holdings in Ethiopiai (m 

1994-1995 1995-1996 Change 1994-1995 1995-1996 

t Source, Central Stati 


tend to be planted in marginal areas without fertilizer. When rains are adequate, 
as in 1995-1996, they can show dramatic increases in production. 

Among the major cereals grown in the country, maize had the largest yield 
gains, with 31%, followed by tef, with 19%. This is to be expected, since these 
two crops use the bulk of the fertilizer in the country. In spite of the success of 
participating farmers in raising yields in the demonstration plots, a significant gap 
still exists between production potential and actual yield among the vast majori- 
ty of small-scale farmers. The challenge ahead is to convert that potential into 


In formulating a fertilizer sector strategy for Africa, it is important to 
understand the nature of industrially produced fertilizer products. Fertilizer man- 
ufacturing units are costly to construct and operate, and they must be relatively 
large in scale and operated near capacity to remain economic. The cost of a typ- 
ical plant that produces 0.5 million t yr- 1 of urea or 0.3 million t yr" 1 of DAP can 
easily exceed U.S. $300 million (Williams & Schultz, 1990). Because the capaci- 
ty of these basic production units is so large, few African countries can justify 
their construction, at least on the basis of domestic demand, even when they have 
the raw materials at hand. Indeed, among African countries, only Nigeria and 
South Africa consume >0. 25 million product tyr 1 (FAO, 1996). 

Africa is endowed with numerous phosphate ore deposits, which are a 
potential source of phosphate fertilizers. With the exception of the export-orient- 
ed facilities of Senegal, Togo, Morocco, and Tunisia, few of these deposits have 
been developed. Some are too small and low in quality for commercial develop- 
ment. But the overriding issue has been limited domestic markets and depressed 
phosphate prices in the global fertilizer market, which do not justify investment 
and operating costs. There are, however, some high-quality deposits that are agro- 
nomically effective, such as in Mali and Tanzania, where simple processing 
would make them suitable for direct application to the soil (Buresh et al., 1997, 
this publication). Use of such deposits should be encouraged. 

In most African countries the demand for mineral fertilizer is low but 
expected to grow appreciably over the next two decades. Thus, the fertilizer sup- 
ply system should be simple but designed for growth in a stepwise fashion to 
meet farmers' needs in a timely way. Schultz and Parish (1989) illustrate a step- 
wise development of a fertilizer supply system to meet growing demand, which 
seems applicable to Africa. Up to about 100 000 tonnes of annual product 
demand, importation is almost certain to be the most cost-effective alternative — 
first in bags and then in bulk. Between 0.1 and 0.2 million tonnes of product 
demand, local production of granular products becomes a potentially viable 
enterprise. Above 0.25 million tonnes of product demand, establishing a domes- 
tic production unit of world class can be considered, if the raw materials such as 
natural gas or phosphate rock are available and if there is an assured market for 
operating the factory at high capacity. 


Meanwhile, African countries can benefit from the more favorable eco- 
nomies of large-scale, export-oriented production units in Europe, the USA, 
North Africa and the Middle East, and elsewhere by importing the required prod- 
ucts. Large benefits can be realized by simply performing the marketing functions 
more efficiently. The landed cost of fertilizer can be decreased by 20 to 40% 
through relatively simple improvements in procurement systems, such as realis- 
tic demand forecasts, selection of appropriate fertilizer types, and consolidation 
of annual fertilizer requirements into large orders to obtain favorable freight on 
board (f.o.b.) prices and ocean shipping rates. Additional savings can be achieved 
by local bagging, local blending of imported materials, and eventually, local 

A cost-effective procurement system should also include unrestricted and 
timely availability of foreign exchange for procuring fertilizer, and a streamlined 
procedure of administrative approval that allows the procurement body to re- 
spond rapidly and take advantage of favorable supply and demand pricing situa- 
tions. Properly valued exchange rates also are important. Continually devaluing 
national currencies work against importing fertilizers and crop protection chemi- 
cals, and therefore against modernization. 

To lower transport costs, there is an urgent need to shift from low- to high- 
analysis fertilizers, as Ethiopia has done. For example, the delivered cost of N to 
African ports can be reduced by about 40% by shifting from ammonium sulfate 
to urea (Williams& Schultz, 1990). There also are additional savings to be real- 
ized on internal freight, warehousing, bagging, and handling. In the Southern 
African Development Community (SADC) countries alone, Donovan (1996) 
reports that US$ 100 million can be saved annually — or about 25% of total out- 
lays — by switching to high-analysis grades and bulk handling. 

Considerable gains also can be made from improving the effectiveness of 
research and extension programs in soil fertility management. More area- and 
crop-specific fertilizer recommendations are needed and, where intensive crop- 
ping is practiced, improved monitoring of secondary nutrients and minor element 
deficiencies. More research also is required to develop integrated practices to 
restore and manage soil fertility, which can involve both inorganic and organic 
nutrient sources. National extension services also need to mount mass-education 
campaigns to teach farmers to use fertilizers in the most efficient manner possi- 
ble, such as appropriate formulas and combinations of nutrient sources, timely 
application, optimum planting densities, and timely weed control. 

Of course, other developments must come to achieve a more effective fer- 
tilizer sector, with improvements in infrastructure high on the priority list. In vir- 
tually all African countries more investments are needed in farm-to-market roads 
to get inputs and outputs in and out of farming areas more efficiently and in trunk 
roads to link the major cities with the main agricultural producing areas. Over the 
longer term, low-cost high-volume transportation systems — such as railroad and 
waterways — must also be given serious consideration. As infrastructure is 
improved, the cost of using modern agricultural technologies will decline signif- 



Plant breeding is the greatest practical achievement of the biological sci- 
ences in the 20th Century. Compared with traditional varieties, today's new food 
crop varieties are vastly more efficient in grain production, in genetic resistance 
to diseases and insects, and in tolerance of various agroclimatic stresses than 
were their predecessor land races. While improved cultivars of maize, rice, 
wheat, barley, sorghum, tubers, pulses, and oilseeds have been developed by agri- 
cultural researchers, too few of these genotypes are reaching African farmers, 
especially small-scale farmers. 

It is very important that African governments look more seriously at the 
issue of developing the seed industry, as expanded use of improved seed is a nec- 
essary condition for a Green Revolution in Africa. As a general principle, seed 
produced by private enterprises is invariably of higher quality than seed produced 
by a public sector organization. Thus, future seed industry strategies should have 
a private sector orientation. But for many crops — especially the food crops grown 
by smallholders — public sector institutions also will have an important role to 
play, especially in crop improvement research to develop new varieties and 
hybrids, in seed certification, and in extension programs to promote rapid diffu- 
sion of these more grain-efficient, fertilizer-responsive genotypes. Restoration of 
soil fertility will help ensure that the potential of improved cultivars is realized. 


Policy planners have failed to fully understand the impact that the intro- 
duction of high-yielding food crop technologies can have on reducing poverty 
and on making agriculture the engine for economic development that it can — and 
must — become in Africa. In most African countries, basic foods are too expen- 
sive rather than too cheap. They are too expensive because of the inefficient, low- 
yielding technologies that are being employed, which neither provide farmers 
with adequate incomes nor consumers with food at accessible prices. Farmers can 
increase their incomes through improvements in productivity and expanded pro- 
duction. More plentiful food supplies lead to lower real prices which, in effect, 
means increased income for all consumers, but with special benefit for the poor, 
since most of their income goes to acquire food. Until this reality is understood, 
and translated into concrete strategies and programs on the ground, the agricul- 
tural growth targets of 5 to 6% needed to revitalize African economies and reduce 
poverty will fall far short of their mark. 

There are still influential environmental groups that advocate reliance on 
organic or natural farming in Africa, on the grounds that N fertilizers are harm- 
ful to the environment, costly to import and expensive for farmers to use, and in 
the long run not sustainable because nonrenewable resources are used to produce 
them. The organic farming advocates, however, fail to recognize that organic 
sources of N fertilizers generally are very limited and that the release of nutrients 
to the soil by natural processes usually do not match the time when crops need 
them. At best, if Asian experience is relevant, reliance on organic fertilizers in 


Africa will give 2% yr" 1 agricultural growth (Hiyami & Ruttan, 1985), well 
below the rates needed to reduce poverty and improve food security. By all 
means, let's use organic sources — and local phosphate resources — where practi- 
cal and cost effective. And let's make sure that sufficient research investments are 
made to understand how best to combine organic and inorganic nutrient sources. 
But let's not fool ourselves — there simply isn't enough organic fertilizer to sup- 
ply sufficient nutrients to the soil to satisfy the growing food demand of Africa. 

Agriculturalists and environmentalists certainly both have a professional 
and moral obligation to warn the political, educational, and religious leaders of 
the world about the magnitude and difficulties of producing ever-greater quanti- 
ties of foods to feed the unrelenting population monster. But by the same token, 
we must face up to the fact that we cannot turn back the clock to the good old 
days of the early 1930s, when the world population stood at two billion people 
and little mineral fertilizer and few agricultural chemicals were used. 

Take the USA as an example. In 1940, the production of the 17 most impor- 
tant food, feed, and fiber crops totaled 252 million tonnes from 129 million 
hectares. Compare these statistics with 1990, when American farmers harvested 
approximately 600 million tonnes from only 119 million hectares — 10 million 
<50 yr previously. If the USA attempted to produce the 1990 harvest with the 
technology that prevailed in 1940, it would have required an additional 188 mil- 
lion hectares of land of similar quality. This theoretically could have been 
achieved either by plowing up 73% of the nation's permanent pastures and range- 
lands or by converting 6 1 % of the forest and woodland area to cropland. In actu- 
ality, since many of these lands are of much lower productive potential than the 
land now in crops, it would have been necessary to convert a much larger per- 
centage of the pasture and rangelands or forests and woodlands to cropland. Had 
this been done, imagine the additional havoc from wind and water erosion, the 
obliteration of forests and extinction of wildlife species through destruction of 
their natural habitats, and the enormous reduction of outdoor recreation opportu- 

Impressive savings in land use have also accrued to China and India 
through the application of modern technology to raise yields (Fig. 4-1). Had the 
cereal yields of 1961 still prevailed in 1992, China would have needed to increase 
its cultivated cereal area by more than threefold and India by about twofold, to 
equal their 1992 harvests. Obviously, such a surplus of agricultural land was not 
available. These lessons of agricultural modernization should not be lost in 

Increasing fertilizer use on food crops from around 5 kg nutrients to 30 to 
40 kg ha" ' of arable land is surely not an environmental problem but a central 
component in Africa's environmental solution. It has been amply demonstrated 
that mineral fertilizers, when properly used, are not only beneficial for produc- 
tivity enhancement but also are environmentally friendly by permitting produc- 
tion on the same land, and thus prevent the migration of farmers to marginal soils 
in search of plant nutrients. Increased fertilizer use also should help to reduce soil 
erosion by increasing plant biomass and vegetative ground cover and, assuming 
that crop residues are returned to the soil, contribute to improving the organic 


■ China, all cereals 

1961 -147 million tonnes 
1992-400 million tonnes. 

1961 1967 1973 1979 1986 1991 

India, all cereals 

1961-87 million tonnes 
1992 - 200 million tonnes 

l i » i Chin nicl Ind in i rniei through raisin i I i I i In upper 

shows the land area needed for 1992 cereal production, had 1961 yields still prevailed: the 
curve shows tiie land area that was actual!) harvested (calculated front FAO production data). 

it of the soil. Obviously, a combination of the two sources of nutri- 
ents — inorganic and organic — is more desirable than the use of one source to the 
exclusion of die other; however, any strategy to achieve 4 to 5% annual agricul- 
tural growth in Africa should have as its central component the increased and 
judicious use of mineral fertilizers, supplemented by organic fertilizer sources; 
not the reverse. 


Birdsall, N. 1995. The coexistence of global food surplus and famine: Poverty's role in the food equa- 
i n. p. 3 4. I ii ii M iii 1 i hi n I i r|n i mm ni t i ni d 

at an Int. Conf, Washington, DC. 13-15 June 199 Int. I I r> h I % , n i 


Borla 1 dial lei till i luction. .1 1 A id A n For. 

Suppl. 21:15-55. 

Borlaug, N.E., and C.R. Dowswell. 1994. Feeding a human population that increasingly crowds a 
ile Planet. Kt note lectin j I Tin >f 15th World mgre if Soil in. 

Acapulco, Mexico. 10-16 July 1994. ISSS, Wageningen, the Netherlands. 

Borlaug, N.E., and C.R. Dowswell. 1995. The importance of agriculture and a stable adequate food 
supply to the II f humankind, p. I I III I se ti ch in domestic 

and international agribusiness management. Vol. i I. JAi Press. Greenwich. C'T. 


Bimib. B. 1995. Growth potential of existing technology is insufficiently tapped, p. 193-205. in Nurul 

I I in I I Poj ul in ii ii i i in h rly twenl first ct nti ' tin i itui food cl in uul 

of an increasing population. Int. Food Policy Res. Inst., Washington, DC. 
Buresh, R.J., PC. Smithson, and D.T. Heliums. 1997. Building soil phosphorus capital in Africa, p. 
111-149. /// R..I. Bu hi tin 1 1 fertility in Afn A Spec. Publ. 51. 

SSSA, Madison, WI (this publication). 
Central Statistical Authority. 1996a. Agricultural sample survey 1995/96: Area and production for 

major crops. Statistical Bull. 152 Vol. 1 . Addis Abaha. Ethiopia. 
Central Statistical Authority. 1996b. Agricultui i mil u y 199 >6 \i ea and production of belg 

season crops. Statistical Bull. 152. Vol. 5. Addis Ababa, Ethiopia. 
Cleaver, K.M., and W.G. Donovan. 1995. Agriculture, poverty and policy reform in sub-Saharan 

Africa. World Bank Dis. Pap. 280. World Bank, Washington, DC. 
' i i I i i i i I I i rh iral: Th pulatioi nculture and envi- 

ronment nexus in sub-Saharan Africa. World Bank, Washington, DC. 
In I)) n ii i In l n i i i ii I ES Work. Pap. 18. 

Africa Tech. Dep. World Bank, Washington, DC. 
Dyson, T. 1995. World food dim. aid uul uppl; pi if In Th F oh/ i Society Proc. 367, 

Cambridge, 6-8 Dec. 1995. Greenhill House. Peterborough. England. 
FAO. 1996. FAO fertilizer yearbook-1995. Vol. 45. Food and Agric. Organization of the United 

Nations, Rome. 
FEWS and EU-LFSU. 1996. Monthly food security bulletin. November 1996. USAID Famine Early 

Warning System and EU Local Food Security Unit. USAID Mission, Addis Ababa, Ethiopia. 
CCA. I '' i ii i ii ultui il i i ii I n Mil in ub I u in Africa. Document 

GCA/EC/No. 2/8/1996. Global Coalition for Africa, Washington, DC. 
Giller, K.E., G. Cadisch, C. Ehaliotis, E. Adams, W.D. Sakala, and PL. Mafongoya. 1997. Building 
il nitrosen capital in Afi 111 I ! I i Replenishh 111 

ity in Africa. SSSA Spec. Publ. 51. SSSA, Madison, WI (this publication). 
Hiyami, Y., and V. Ruttan. 1985. Agricultural development: An international perspective. John 

Hopkins Univ. Press, Baltimore, MD. 
' in n I I i 1 uli i ml welfare d I ment. The Fertili iety Pri in i ! 

Greenhill House. Peterborough, England. 
Rosegrant, W.M., M. Agcaoli-Sombilla, and N.D. Perez. 1995. Global food projections to 2020: 

Implications for investments. Dis. Pap. 5. Int. Food Policy Res. Inst., Washington, DC. 
Sanchez, P.A., K.D. Shepherd, M.J. Soule, F.M. Place, A.U. Mokwunye, R.J. Buresh, F.R. Kwesiga, 

A.-M.N. Izac, C.G. Ndiritu, and PL. Woomer. 1997. Soil fertili! i pleni hment in Africa An 
11 II 11 1 II hi I [ I ll | I / i I i I I r I i I I 

tility in Africa. SSSA Spec. Publ. 51. SSSA, Madison, WI (this publication). 
Schultz. J.. I.. .Hid D.H. Parish. 1989. Fertilizer production and supply constraints and options in sub- 
Saharan Africa. Pap. Ser. P-10. Int. Fert. Dev. Ctr., Muscle Shoals. AL. 
Smaling, E.M.A., S.M. Nandwa, and B.H. Janssen. Soil fertility in Africa Is at stake, p. 47-61. In 

R.J. Buresh et al. (ed.) Replenishing soil fertility in Africa. SSSA Spec. Publ. 51. SSSA, 

Madison, WI (this publication). 
i I. J i i i I ii iin I ' i \ m ii i i 1 mi i tion in sub Sahaian 

Africa, 1983-2000. Rep. 28. DLO Winand Staring Ctr. for Integrated Land, Soil and Water 

Res., Wageningen, the Netherlands. 
Williams, L.B., and J.J. Schunz I'm rtilizer suppl trat i i i Saharan Africa, p. 

I I / 1 i i il h i i i 1 uii 1 1 i lin H I | i in ii 

Africa. 1989. Sasakawa Africa Assoc. Mexico City. 

Soil Profile Alteration under 
Long-Term, High-Input 

Stanley W. Buol and Michael L. Stokes 

North Carolina State University 
Raleigh, North Carolina 


This chapter presents the findings of several studies that suggest beneficial soil 
property and characteristic changes from long-term, high-input agronomic methodology. 
Organic C content declined progressively in unfertilized continuous maize (Zea mays L.) 
plots on Mollisols in the Morrow plots in Illinois until complete fertilization was practiced 
in 1955. Organic C content in the previously mentioned unfertilized continuous maize 
plots increased significantly during an 18-yr period after applying lime, N, P, and K at soil 
test recommended rates. In a Canadian study on Aquolls, soil organic C content was high- 
er after 32 yr of continuous maize in fertilized plots than in similar nonfertilized plots. Two 
separate studies invi ti ting the long-term effect of erosion on the productive potential of 
several Mollisols and Ultisols both predict only nominal losses in productivity as a direct 
result of erosion. Long-term, high- input management of Ultisols in North Carolina signif- 
icantly increased exchangeable basic cation levels with corresponding decreases in 
extractable Al levels in subsoil horizons; however, no apparent downward translocation of 
P was detected. These experiences with high-input agricultural practices in North America 
can provide an understanding of the long-term consequences and sustainability that such 
practices could have on similar soils in Africa. 

Intensive crop production with modern technology has altered many soils so that 
they are now more productive. Clearly high-input agriculture has increased the 
food production per unit of land in the USA since the middle of the 20th centu- 
ry. As seen in Fig. 5- 1 , chemical fertilization has been one of the factors that have 
increased efficiency of land area for food production. Concomitant with increased 
fertilizer use has been mechanization of tillage and harvest operations, improved 
food storage and distribution, and improved genetics. These practices rely on the 
availability of raw materials and infrastructure; tints their desirability and sus- 
tainability can be debated with regard to social and economic parameters. In this 
chapter, we address what impacts these practices, in sum termed high-input agri- 
culture have had on soil properties. From this review of experience with high- 


Total hectare* harvested 

U.S. populations^ 

\ Hectares harvested lor 
\ U.S. consumption 

\3SSWm * 

10 6 tonnes fertilizer^ 6 

5.0 6j5 ^^ ^^ ^m 

1910 1920 1330 1940 1350 1960 1370 TWO 1330 


Fig. 5- 1 . Cropland harvested, fertilizer use, and population in the U.S., 1910-1990 (after Buol, 1995). 

input agriculture in North America, scientists in Africa may better understand the 
long-term consequences and potential sustainability of such practices. The results 
reported on the long-term effect of lime and fertilizer use on Ultisols that at one 
time were subjected to slash-and-burn management should have direct applica- 
tion to the evaluation of large areas of Africa with similar acid, low-activity clay 


One of the best records of continuous crop production in the USA comes 
from the Morrow plots on the University of Illinois campus at Urbana- 
Champaign. The plots were established in 1876 on nearly level Flanagan silt 
loam (fine, smectitic, mesic Aquic Argiudoll) soils (Odell et al., 1982). There was 
no record of the organic C content of the site when the plots were established, but 
four samplings of the plots in 1904, 1913, 1923, and 1933 were reported by 
DeTurk (1938; Table 5-1). He noted that the organic C content progressively 
declined in the nonfertilized continuous maize, the nonfertilized maize-oat 
(Avena sativa L.) rotation, and the fertilized, limed, and manured continuous 
maize plots, but did not appear to decline in the maize-oat-clover (Melilotus alba 
L.) rotation and the fertilized, limed, and manured maize-oat rotation plots. 

In 1944 the surface soil under the sod around the plots was sampled and 
compared with the surface soil in the then 68-yr-old plots (Table 5-2). Odell et 
al. (1982) reported that organic C and total N contents in the continuous, nonfer- 
tilized maize plots were lower than in the borders. The limed, manured, and P fer- 
tilized maize-oat-clover rotation plots, where the clover crop was not harvested 


No fertilizer MLP§ 

Source, DeTurk(1938). 

Method of organic C analysis not given and ma 

MLP. manure, limestone, ami phosphate rock. 

tents nearly the same as the plot borders. In 1955, the continuous unfertilized 
maize plots were split and lime, N, P, and K were applied at rates corresponding 
to soil test recommendations. This resulted in a significant increase in soil organ- 
ic C content between 1955 and 1973 even with continuous maize harvest (Fig. 
5-2). Darmody and Norton (1994) concluded from micromorphological studies 
of the Morrow plots that fertilizer and lime inputs had little effect on aggregate 
properties or soil fabric; however, aggregate stability and macro-pore continuity 
decreased as intensity of cultivation increased. 

Gregorich et al. (1996) found higher organic C contents to a depth of 0.26 
m in the fertilized plots after 32 yr of continuous maize with and without N, P, 
and K fertilization on fine, mixed, mesic Typic Haplaquolls in Ontario, Canada. 
The difference was accounted for by more Q-derived C in the fertilized soils. 
They concluded that adequate fertilization not only increased crop yields leading 
to greater C storage in the soil but did not significantly alter the rate of native soil 
organic C turnover as determined by 13 C studies. 

Although a decline in soil organic C content upon cultivation is well docu- 
mented, there is scant evidence that lower soil organic C contents have any 
adverse effect on the production potential of the soil. Numerous publications have 
attributed a decrease in available water-holding capacity to lower soil organic C 
content. This notion is in part created by the fact that bulk density increases as 
soil organic C content declines, and cultivation practices tend to compact the sur- 

Table 5-2. Organic n 

it of the surface soil in and around the Morrow plot; 

Source, Odell et al. (1982). 

'., percentage of sod bonier least). 

MLP, manure, limestone, and phosphate re 



1903 1913 1923 1933 1943 1953 1963 1973 

RC = Regression coefficient (t ha -1 in plow layer) 

* ■ Significant at 0.05 level; " = Significant at 0.01 level 

rr content of the plow layer in north blocks of Mor 

face horizons. When available water capacity is reported on a soil-weight basis, 
values increase with increased soil organic C content. When measured on a soil- 
volume basis, this relationship does not exist. Since plants root in a volume of soil 
rather than a weight of soil, the volumetric relationship is to be preferred in eval- 
uating the production potential of a soil. Bauer and Black (1992) studied the 
effect on management-induced changes in soil organic C concentration on the 
available water capacity of three soil textural groups in North Dakota. Forty-eight 
sampling locations representing different soil-management systems were select- 
ed and sampled in four increments to a depth 0.46 m. All of the soils were 
Haploborolls or Argiborolls in coarse-loamy, fine-loamy, fine-silty, or fine parti- 
cle-size families and mixed or smectitic mineralogy families. The bulk density, 
soil organic C content, and available-water capacity of each sample were then 
determined, and regression analysis of the data was conducted and presented by 
particle-size group (Table 5-3). In the medium and fine particle-size samples, 
available-water capacity, on a volume basis, decreased as soil organic C contents 
increased. In the sandy soils, available-water content was unaffected by organic 
C content. They concluded that any decrease in production potential resulting 
from a decrease in soil organic C content associated with management could not 
be caused by decreased available-water capacity. Additional studies (Bauer & 
Black, 1994) reported that decreased crop productivity associated with soil 
organic C contents resulted from a concomitant loss in fertility, primarily miner- 
alizable N. 

Decomposing organic residue does have a positive affect on aggregate sta- 
bility and infiltration. Bruce et al. (1995) determined that approximately 12 t ha" 1 
yr- 1 of decomposing mulch was required on the soil surface to significantly 
increase aggregation and infiltration of eroded Udults in the thermic soil temper- 


Table 5-3. Effect of soil organic C concentration on bulk density and available water capacity of threi 
soil textural groups calculated from regressions. 

t Source, Bauer and Black (1992). 

j BD, bulk density. 

§ AWC, available water capacity. 

ature regimes of Georgia. This effect appears related to the binding properties of 
fungi exudates. Fungi are especially active in decomposing organic residue. The 
quantity of crop plant residue added to the soil appears to be the most significant 
organic factor in developing and maintaining a stable surface soil structure 
(Gantzer et al., 1987); however, it follows that higher rates of organic residue 
return result in higher quantities of soil organic C if all of the other parameters of 
soil temperature and aeration are equal. 


"Under the natural vegetation, especially in humid forested regions, natur- 
al erosion serves to maintain a degree of youthfulness and fertility in many soils 
by removing leached materials from the soil surface while new materials are 
added to the soil profile from beneath" (Soil Survey Staff, 1951, p. 251). 
Subsistence farmers without access to fertilizers often concentrate on sloping 
land, attempting to obtain meager yields from the natural weathering of minerals 
exposed as surface horizons rapidly erode. Erosion is accelerated as they deplete 
the fertility by harvesting food products and "if continuing cultivation is 
assumed, generally, although not always low soil fertility can be regarded as a 
main cause of erosion" (Soil Survey Staff, 1951, p. 261). 

There is no doubt that the intense cultivation and monoculture cropping 
systems of high-input agriculture also can accelerate erosion rates. The distinc- 
tion between natural erosion and accelerated erosion is not easy to determine on 
every soil (Soil Survey Division Staff, 1993). Obviously areas cannot be classi- 
fied on the basis of what is gone — no longer there. "The mapping standards and 
guides must be based on characteristics that can be seen" (Soil Survey Staff, 
1951, p. 260). "Erosion by itself, unrelated to the [type of] soil, means little or 
nothing. Tons or inches of soil lost through erosion have little general meaning in 
terms of soil productivity" (Soil Survey Staff, 1951, p. 268). While amount of soil 
loss is important to downstream sedimentation, the impact of accelerated erosion, 
it of the total erosion resulting from the cultivation of land, on crop 


Table 5-4. Change in production index h\ slope class for sods in Major Land Resource A 
Minnesota after 25, 50, and 100 yr of erosion.t 

100 yr 


56 700 


190 755 


91 125 


22 680 


8 505 

Source, Pierce e 

■t al. (19S 


production potential must be related to the soil properties remaining after the ero- 
sion has taken place. Clearly, each kind of soil is affected differently by erosion. 

Pierce et al. (1983) addressed this question by considering the long-term 
effects of erosion on the productive potential for maize on some Mollisols in 
Minnesota. They assumed that the soil received good management, which includ- 
ed fertilizer replacement of elements removed in harvest. They considered only 
the loss in production potential resulting from physical changes resulting in a 1- 
m-deep root zone as material was eroded from the soil surface at rates currently 
reported for the respective soils. For Monona soils (fine-silty, mixed, mesic Typic 
Hapludolis) they projected that an annual erosion rate of 76 t ha- 1 yr"' would 
result in a production index reduction of 3% after 100 yr of simulated erosion. 
Kenyon soils (fine-loamy, mixed, mesic Typic Hapludolis), eroding at a rate of 
17 t ha" 1 yr- 1 would have a production index reduction of 4% after 100 yr. 
Rockton soils (fine-loamy, mixed, mesic Typic Argiudolls) would be expected to 
have a 20% reduction in production index in 100 yr with an erosion rate of 25 t 
ha"' yr- 1 . The greater influence of erosion in the Rockton soils results from 
decreasing the depth to a less desirable root environment in the argillic horizon 
present in Argiudolls and not present in the more uniform distribution of particle- 
size with depth in the Hapludolis. When they applied 100-yr projections to the 
367 939 ha of cropland in Major Land Resource Area 105 (MLRA 105) in 
Minnesota they predicted the production index loss, by slope class, after 25, 50, 
and 100 yr at the present erosion rates (Table 5-4). Clearly the greatest potential 
loss of production is on the 8.4% of the cropped land with slopes greater than 

The results of Pierce et al. (1983) reflect potential productivity loss on what 
are considered some of the best soils in the world. The fact that the soils studied 
are formed in thick loess and glacial till deposits accounts for the small amount 
of physical change that could be expected as surface soil is eroded and the root 
zone projected into subsoil material. It could be expected that erosion would 
more severely decrease the production potential of Ultisols where finer textured 
argillic and kandic horizons of low base saturation are exposed by erosion of top- 

Daniels et al. (1989) intensively investigated crop productivity in fields in 
the piedmont of North Carolina to determine if the soil properties considered to 


result from the process of erosion affected crop productivity. The fields they stud- 
ied had been in cultivation since 1800 or earlier. Although a complete record is 
not possible, it is probable that the original cultivation was as shifting cultivation, 
progressing with time to cultivation with animal-powered machines and today 
conducted by mechanized farming practices, with lime and fertilizer applications 
becoming common practice prior to 1900. Soils included in the fields studied 
were classified as Cecil and Georgeville (clayey, kaolinitic, thermic Typic 
Kanhapludults and Typic Hapludults, respectively) and Cullen and Vance (clayey, 
mixed, thermic Typic Hapludults). These soils are formed in saprolite weathered 
from granite, gneiss, and other felsic rocks. Each of the 16 fields studied was tra- 
versed while in a vegetation-free state prior to planting, and plots that represent- 
ed defined erosion classes and landscape position were selected and surveyed rel- 
ative to identifiable locations. In this way the researchers could identify and com- 
pare combinations of soil properties considered to result from erosion and known 
to influence crop production within comparable landscape positions. The land- 
scape positions identified were linear slopes, nose slopes (convex), head slopes 
(concave), foot slopes, interfluves, and shoulders. Erosion classes were defined 
as slight, moderate, or severe according to the amount of B horizon (argillic or 
kandic) material incorporated into the plow layer. When yields from various 
crops during a span of 5 yr were summarized they found that soil features con- 
sidered to be a result of moderate and severe erosion resulted in an economic loss 
of only US$ 4.44 ha" 1 yr 1 based on 1987 prices for crops produced. Although 
distinct production differences were present among all plots studied, most of 
these differences were related to landscape position, not to soil property differ- 
ences attributed to erosion within the same landscape positions. Daniels et al. 
(1989) clearly found that slightly eroded soils usually out yielded severely erod- 
ed soils, but the differences were small and often confounded as a result of weath- 
er patterns that exaggerated differences some years, even reversing them in some 
years. "Because the area of severely eroded soils in most fields is small, their 
impact on field production is slight" (Daniels et al, 1989). 

These results question whether the features usually attributed to erosion 
have really resulted because of accelerated erosion or whether they are at least in 
part the result of pedogenesis related to the topographic or relief factor of soil for- 
mation (Jenny, 1941). When McCracken et al. (1989) examined a 12-ha tract of 
virgin land in the piedmont area of North Carolina, with Cecil soils and slopes 
ranging from 2 to 10%, they found 60% of the area would have considerable 
amounts of the upper B horizon incorporated into an Ap horizon should the area 
be plowed. Since incorporation of B-horizon material into the Ap horizon is taken 
as evidence of erosion, these soils would be identified as borderline moderately 
eroded even though they had never been cultivated. 

Erosion, whether associated with subsistence or intensive agriculture, can 
severely decrease the production capacity of some soils; however, the experi- 
ences cited indicate that degradation of production potential from erosion under 
high-input management is probably quite modest except on sloping parts of the 
landscape that have subsoils or parent rock that is not capable of supporting root 
growth. Unqualified observations indicate that, with the advent of high-input 
agriculture, steep slopes most subject to severe erosion are often not cultivated 


because of their incompatibility with the use of large machines so intrinsic to 
high-input farming. Also, if larger crop plants result from a high input of fertiliz- 
er, greater amounts of root and straw residue are produced. It is the decomposi- 
tion of organic residue that has positive affects on increased aggregation and infil- 
tration (Bruce et al., 1995; Gantzeretal., 1987). 


Intensive crop production on naturally infertile, acid Ultisols requires inten- 
sive application of lime and fertilizers to establish an adequate medium for high 
yields from crop plants, and then to replace nutrients removed at. harvest. 
Historical records indicate that lime and fertilizers have been used on Ultisols in 
North Carolina for > 100 yr. Prior to general use of lime and fertilizer, shifting cul- 
tivation was extensive in this state. A survey was conducted in North Carolina to 
ascertain changes in the chemical status of Ultisol profiles as a result of long-term 
intensive application of lime and fertilizer (M.L. Stokes, 1995, unpublished data). 

Three soil cores to a depth of 2 m were taken each from directly adjoining 
managed and unmanaged areas of a soil map unit. Each core was divided into 
separate constituent morphological horizons for analysis. The distance between 
managed and unmanaged sampling points was kept to a maximum of 50 m to 
minimize spatial variability. Sampled map units were investigated as thoroughly 
as conditions allowed to confirm that the managed sections had been limed and 
fertilized for at least 20 yr. The sites reported in this article had been verifiably 
limed and fertilized as required to obtain maximum potential crop yields for >30 
yr. The adjacent unmanaged areas were determined to have not been fertilized for 
>50 yr by the estimated ages of existing trees. Samples from each horizon were 
processed for exchangeable base cations, exchangeable Al, soil pH, available P, 
and total P. Exchangeable base cations were extracted with 1 A/NH 4 OAc at pH 
7.0 (Soil Survey Laboratory Staff, 1992, p. 146-148). Exchangeable Al was 
extracted with 1 M KC1 (Soil Survey Laboratory Staff, 1992, p. 193-196). 
Available P was determined by the Bray I extraction method (Soil Survey 
Laboratory Staff, 1992, p. 218-220). Total P was determined by complete diges- 
tion in concentrated H2SO4 and H2O2. 

The results presented here are for two soils, a Norfolk sandy loam (fine- 
loamy, siliceous, thermic Typic Kandiudults) from the coastal plain in North 
Carolina and a Cecil sandy loam (clayey, kaolinitic, thermic Typic 
Kanhapludults) from the piedmont in North Carolina. Average yields on Norfolk 
soils receiving a high level of management are reported as being 6.4 t ha' 1 for 
maize and 2.21 ha" 1 for soybean [Glycine max (L.) Merr.; Shaffer, 1994]. Average 
yields on Cecil soils with similar management are reported as being 6.01 ha" 1 for 
maize and 2.2 t ha" 1 for soybean (Spangler, 1994). 

The exchangeable cation extractions, considered from a percentage of sat- 
uration basis, indicate a marked increase in percentage of base saturation [(cmol c 
Ca + Mg + K)/(cmol c Ca + Mg + K + Al) x 100] in the managed sites of both the 
Cecil and in the Norfolk soils to a depth of 2 m. There is also a decrease in per- 
centage of Al saturation [(cmol c Al)/(cmol c Ca + Mg + K + Al) x 100] in the man- 
aged sites of both soils (Fig. 5-3 and 5-4). Calcium was the major exchangeable 


Al saturation (%) 

30 40 50 60 70 80 90 10' 

Fig. 5-3. Effect of management on Al saturation for a Cecil sandy loam (2 to 6% slope, eroded, 
clayey, kaolinitic. thermic Typic Kanhapludultsi Data points represent subsampling from entire 
horizon; error bars represent the standard error of the mean (P = 0.05). 

Al saturation r%) 
10 20 30 40 50 60 70 80 90 100 




'-*- Managed 
■■•■■ Unmanaged 

1 — 1 II I 

25 to 50 
50 to 75 
75 to 100 
100 to 125 
125 to 150 
150 to 175 
175 to 200 

Fig. 5-4. Effect of management on Al saturation tor a Norfolk sandy loam (2 to 6<< slope, fine-loamy, 
siliceous, thermic Typic Kandiudults). Data points represent subsampling from entire horizon: error 
bars represent the standard error of the mean (P • 0.05). 


Exchangeable Ca (cmol c kg" ' soil) 

Fig. 5-5. Effect of management on exchangeable Ca for a Cecil sancK loam (2 to 6% slope, ere 
I I In ii il inn 1 | in j lucl i' 1 1 | i i i i i nt snl mplin in e 

horizon; error bars represent the standard error of the mean IP = 0.05 1. 

Exchangeable Ca (cmol c kg" 1 soil) 
0.2 0.4 0.6 0.8 t.O 

-*- Managed 
■■•■■ Unmanaged 

Fig. 5-6. Effect of management on exchangeable Ca for a Norfolk sand) loam (2 to 6% slope, fine- 
loamy lliceous thermic 1 Kandiudults). D i H hom entire hori- 
zon; error bars represent the standard error of the mean (/' = 0.05). 


base cation component of the base saturation levels in all soils and was observed 
to have a dramatically greater accumulation in managed sites (Fig. 5-5 and 5-6) 
from the surface to a depth of 2 m. The apparent increase in the surface horizon 
of the Norfolk (Fig. 5-6) is due to an uncharacteristically large amount of Ca in 
the surface horizon of one of the cores from the unmanaged site as evidenced by 
the large error bar. Exchangeable Mg contents were found to mimic exchangeable 
Ca trends in both soils. Dolomite is the most commonly used liming material. 

Exchangeable K contents had little to no observable trends in either man- 
aged or unmanaged soils. Extractable Na only appears in trace amounts in these 

Total P contents were greater in the surface horizons of the managed sites 
reflecting long-term accumulation of fertilizer P (Fig. 5-7 and 5-8); however, 
from 0.75 to 1.5 m in the Norfolk and from 0.5 to 0.75 and 1.25 to 1.5 m in the 
Cecil, the levels of total P were lower in the cultivated sites than in the unman- 
aged sites. We hypothesize that this has resulted from the extraction of subsoil P 
by crop roots active at these depths as a result of diminished levels of extractable 
Al. Available P contents below the surface horizons in both managed and unman- 
aged sites were too low to report effectively. The only substantial differences 
between the two soils was in total amounts of extractable constituents with 
greater amounts extracted from the clayey, kaolinitic soils than from the fine- 
loamy, siliceous soils. A reduction in the adverse effects of short-term drought 
could be expected to result as plant roots are exposed to more available water as 
the effective rooting depth deepens from the reduction in extractable Al, but we 
have no measurements to verify this effect. 

Total phosphorus (mg kg' 1 soil) 

Fig. 5-7. Effect of management on total soil P tor a Cei il sandy loam (2 io (>'', slope, eroded, clayey, 
kaolinitic. thermic Typic Kanhaplndnltsi. Data points represent subsampling from entire horizon; 
error bars represent the standard error of the mean IP = ().():>i. 


Total phosphorus (mg kg" 1 soil) 

Fiji. 5-8. Effect ot'iminajienient on total soil P for a Norfolk sandy loam i ' to ;>'< slope, fine-loamy, 
siliceous, thermic Typic Kandiudults). Data points represent subsampiinji from entire horizon; error 
bars represent the standard error of the mean (P = 0.05). 


High-input agriculture allows agricultural concentration on the soils least 
subject to degradation where, with replenishment of nutrients harvested, produc- 
tion is sustainable indefinitely (Loomis, 1984). While Loomis reached this con- 
clusion from data derived in large part from fertile Mollisols and Alfisols, our 
more recent work clearly indicates substantial improvement in soil properties 
related to crop production resulting from long-term, high-input management of 
naturally infertile Ultisols. High erosion rates are a cause for concern because 
agricultural production potential is determined by the quality of the remaining 
soil. Organic C contents that become lower under inadequate fertilization appear 
to recover when adequate fertilizer is applied. Adequate fertilization also con- 
tributes to greater biomass production tending to protect soil from erosion and 
providing greater quantities of residue critical to soil aggregation. We therefore 
conclude that long-term, high-input agriculture has had a strong positive affect in 
improving agronomic properties of soils. 


Bauer, A., and A.L. Black. 1992. Organic C effects on available w 

groups. Soil Sci. Soc. Am. J. 56:248-254. 
Bauer, A., and A.L. Black. 1994. Quantification of the effect of sc 

productivity. Soil Sci. Soc. Am. J. 58:185-193. 


Brace, R.R., G.W. Langdale, L.T. West, and W.P. Miller. 1995. Surface soil degradation and soil pro- 
ductivity restoration and maintenance. Soil Sci. Soc. Am. J. 59:654-660. 

Buol, S.W. 1995. Sustainability of soil use. Annu. Rev. Ecol. Syst. 26:25 A 4. 

lu 1 ^ i ! 1 ' < ii n ill [ill 1 u ■ ploi wh Hi omj nied long-continued 

cropping. Soil Sci. Soc. Am. Proc. 3:83-85. 

Daniels, R.B., J.W. Gilliam, D.K. Cassel, and L.A. Nelson. 1989. Soil erosion has limited effect on 
field scale crop productivity in the Southern Piedmont. Soil Sci. Soc. Am. J. 53:917-920. 

Darmody. R.G.. am! L.I). Norton 1994. Structural degradation of a prairie soil from long-term man- 
agement. Dev. Soil Sci. 22:641-649. 

Gantzer, C.J., G.A. Buyanovsky, E.E. Alberts, and PA. Remley. 1987. Effects of soybean and corn 
residue decomposition on soil strength and splash detachment. Soil Sci. Soc. Am. J. 

Gregorich, E.G., B.H. Ellert, C.F. Drary, and B.C. Liang. 1996. Fertilization effects on soil organic 
matter turnover and corn residue C storage. Soil Sci. Soc. Am. J. 60:472 A 176. 

Jenny, H. 1941. Factors of soil formation. McGraw-Hill Book Co., New York. 

Loomis, R.S. 1984. Traditional agriculture in America. Annu. Rev. Ecol. Syst. 15:449-478. 

McCracken, R.J., R.B. Daniels, and W.E. Fulcher. 1989. Undisturbed soils, landscapes, and vegeta- 
tion in a North Carolina piedmont virgin forest. Soil Sci. Soc. Am. J. 53:1146-1152. 

Odell, R.T., W.M. Walker, L.V. Boone, and M.G. Oldham. 1982. The Morrow plots: A century of 
learning. Agric. Exp. Stn. Bull. 775. Univ. of Illinois. I'rbana-Champaign. 

Pierce, F.J., W.E. Larson, R.H. Dowdy, and W.A.P. Graham. 1983. Productivity of soils: Assessing 
in i mi u i lue to erosion. J i atei Conserv. 38:39-44. 

Shaffer, K.A., 1994. Soil survey of Northampton County, North Carolina. USDA-SCS. U.S. Gov. 
Print. Office, Washington, DC. 

Spangler, D.G. 1994. Soil survey of Harnett County, North Carolina. USDA-SCS. U.S. Gov. Print. 
Office, Washington, DC. 

Soil Survey Division Staff. 1993. Soil survey manual. USDA Agric. Handb. 18. U.S. Gov. Print. 
Office, Washington, DC. 

Soil Surve\ Laboratory Staff. 1992. Soil survey laboraton methods manual. Soil Surv. Invest. Rep. 
42. Version 2.0. USDA-SCS-NSSC. U.S. Gov. Print. Office, Washington, DC. 

Soil Survey Staff. 1951. Soil survey manual. USDA Agric. Handb. 18. U.S. Gov. Print. Office, 
Washington, DC. 


Building Soil Phosphorus 
Capital in Africa 

Roland J. Buresh and Paul C. Smithson 

International Centre for Research in Agroforestry 
Nairobi, Kenya 

Deborah T. Heliums 

International Fertilizer Development Center 
Muscle Shoals, Alabama 


Cropping with little or no P inputs has depleted soil P stocks in sub-Saharan Africa. 
Phosphorus is now a limiting nutrient in many sandy soils of the semiarid tropics and in 
acid, weathered soils of the subhumid and humid tropics. Replenishing P stocks in highly 
P-deficient soils requires input of P fertilizers rather than sole dependence on P cycling 
through organic-based systems. The soil P stock that serves as a major sink for added P 
and gradually tele |l nl ilabl PI r up to h i i r ired to as soil P capital. It is 
comprised of inorganic P sorbed on clays and Fe and Al oxides and P in organic pools. 
Soil P can be replenished with either soluble P fertilizers, direct application of sufficient- 
ly reactive phosphate rock (PR), or the combination of soluble P fertilizer and PR. 
Phosphorus can be replenished either immediately with a large, one-time P application or 
gradually with moderate seasonal applications of P at rates sufficient to increase avail- 
ability of soil P. The P rate required to overcome P deficiency increases with increasing P- 
sorption capacity of the soil, and the i lual lefil t P ii ises with increas- 

ing P sorption, except for soils with allophane. An initial, large corrective P application 
followed by maintenance applications of P can rapidly replenish and maintain soil P cap- 
ital. The replenishment and subsequent maintenance of soil P must be accompanied with 
management to overcome other nutrient limitations and crop growth c 
sustained crop production at increased levels without environmental d 

Many soils in sub-Saharan Africa (hereafter referred to as Africa) are character- 
ized by deficient levels of plant-available P. Despite diversity in distribution of 
parent material and conditions affecting soil formation, soil P deficiencies pri- 
marily result from either inherent low levels of soil P or depletion of soil P. In 


much of the semiarid zone, soils were derived from acidic parent material that 
contained low levels of P. For the once P fertile soils (e.g., soils of the highlands 
of East Africa), on the other hand, soil P stocks have decreased as increasing pop- 
ulation has led to replacement of traditional systems of shifting cultivation with 
shorter duration unsustainable fallow systems and sedentary agriculture. Many 
smallholder farmers have lacked the financial resources to purchase sufficient 
fertilizers to either correct inherent low levels of soil P or replace the P exported 
with harvested products (World Bank, 1994; Sanchez et al., 1997, this publica- 
tion). This factor has been a major component of the overall depletion of soil fer- 
tility and the resulting degradation of soil resources. 

There is an indisputable need to correct deficiency of soil P in Africa 
(World Bank, 1994; Mokwunye et al., 1996; Smaling et al., 1997, this publica- 
tion). Numerous studies in Africa have shown that P fertilizers — including 
ground PR, modified PR products, and soluble sources, such as triple superphos- 
phate (TSP) and single superphosphate (SSP) — can significantly increase crop 
yields (Wild, 1973; Le Mare, 1984; Sale & Mokwunye, 1993). Relatively small 
seasonal applications of soluble P fertilizer can mitigate P deficiency in soils with 
low to moderate P-sorption capacity; but larger rates of soluble P are required for 
soils with higher P-sorption capacity. 

In view of this situation, there has been a call for the replenishment of soil 
fertility as a capital investment in Africa (World Bank, 1994; Sanchez et al, 1996, 
1997, this publication). The replenishment of soil P fertility implies building soil 
P stocks to levels at which P is not limiting. After P is added to soil, it can be 
sorbed onto clay and oxide surfaces or converted to organic P and then gradual- 
ly released as plant-available P for a number of years. The stock of soil P that 
gradually supplies plant-available P to crops for about 5 to 10 yr has been referred 
to as P capital (Sanchez & Palm, 1996). 

The replenishment of soil P capital can be achieved either rapidly through 
a one-time investment in a large P application or gradually through seasonal 
applications of P at rates sufficient to increase P availability. The replenishment 
of soil P capital, in combination with management to overcome other nutrient 
limitations and crop growth constraints, would provide benefits of increased crop 
production and income to farmers as well as environmental benefits to society 
(Izac, 1997, this publication). After replenishment of soil P with a large, one-time 
investment, only moderate inputs of P would be required to maintain P fertility. 
The combination of P replenishment with N replenishment can have synergism. 
For example, the elimination of P deficiency can enhance N? fixation by legumes 
(Cassman et al, 1993; Giller et al., 1997, this publication), and the integrated use 
of P fertilizers with organic materials to supply N and K can potentially enhance 
P availability (Palm et al, 1997, this publication). 

This chapter will review (i) soil P pools and transformations in relation to 
the concept of P capital, (ii) sources of P including indigenous PR, (iii) the poten- 
tial of organic materials and agroforestry to supply P, (iii) the management and 
economics of P sources, and (iv) conditions conducive for investment in the rapid 
replenishment of soil P. The review will be limited to annual food crops. 



Inorganic Phosphorus 

Phosphorus occurs in soil in inorganic and organic forms, which vary in 
their rate of P release (Fig. 6-1). The original soil source of soluble inorganic P 
(Pj) is dissolution of primary P minerals, mainly apatite. Primary P minerals 
decrease in soil with increasing soil weathering and are relatively unimportant in 
highly weathered soils (Smeck, 1985). Once in the soil solution, Pi can be (i) 
taken up by plants, (ii) taken up by soil biota and converted to organic P (P ), and 
(iii) sorbed onto soil minerals. Labile P; is loosely sorbed and in rapid equilibri- 
um with soil solution Pj, whereas strongly sorbed Pj is only slowly released and 
made available to plants. 

In acid soils without allophane, the sorption of Pj occurs mainly on hydrous 
oxides of Fe and Al at the surface of layer-silicate clay particles (Schwertmann 
& Herbillon, 1992; Frossard et al., 1995), which increase in importance in soils 
with increased weathering and clay content (Sanchez & Uehara, 1980). 
Numerous studies have shown that P sorption is directly related to the contents of 
clay and oxides of Fe and Al in soil (Juo & Fox, 1977; Loganathan et al., 1987; 
Owusu-Bennoah & Acquaye, 1989). Phosphorus sorption is a dominant process 
controlling P availability in Ultisols and Oxisols with medium- and fine-textured 
topsoil (Sanchez & Uehara, 1980), which are major soils in the humid and sub- 
humid tropics of Africa (Deckers, 1993). Phosphorus sorption is of minor impor- 
tance in sandy soils that are predominant in the semiarid tropics of Africa 
(Doumbia et al., 1992). Volcanic soils with amorphous alumino-silicates and 
humus-Al complexes have a very high capacity to sorb P. This sorption tends to 

Fig. 6-1 . Conceptual diagram of soil P pools and P dynamics following applies 
and soluble P fertilizer (adapted from Stewart & Shajpley, 1987). 


be essentially irreversible, and even high rates of P application do not satisfy the 
P-sorption capacity (Espinosa, 1992). 

Inorganic P along the continuum of decreasing rate of release to solution P ; 
(Fig. 6-1) can be distinguished through sequential P fractionation with resin, 
sodium bicarbonate, sodium hydroxide (NaOH), and hydrochloric acid (HC1; 
Hedley et al, 1982; Tiessen & Moir, 1993). Resin-extractable Pi is directly 
exchangeable with the soil solution, and bicarbonate-extractable Pi is loosely 
sorbed. Both fractions represent labile Pj. NaOH-extractable Pi is strongly sorbed 
onto Fe and Al oxides and clay edges and is only slowly available to plants. 
Dilute HCl-extractable Pi represents P from calcium phosphates, which comprise 
a large portion of total Pj on young soils but are relatively unimportant on highly 
weathered soils unless these soils are amended with PR (Tiessen & Moir, 1993; 
Cross & Schlesinger, 1995). 

The NaOH Pj fraction increases in importance with increased soil weather- 
ing due to increased Fe and Al oxides and sorption of P. NaOH P : was the domi- 
nant fraction related to availability of P to plants in an 18-yr continuously culti- 
vated and fertilized cropping system on an Ultisol (Beck & Sanchez, 1994). 
NaOH Pi appears to include a slowly labile Pi pool, which is an important source 
of available P within the time frame of 2 to 10 yr. Studies on Ultisols fertilized 
with P for 4 yr (Linquist et al., 1997a) and 10 yr (Schmidt et al., 1996) found that 
the bicarbonate Pi and NaOH Pi fractions were sinks for fertilizer P applied in 
excess of plant uptake. The amounts of P in these fractions decreased after dis- 
continuation of P fertilization, indicating that the fractions represented sources of 
plant-available P. 

Organic Phosphorus 

Inorganic P in soil solution is partly replenished by mineralization of P . 
Total Po decreases with continuous cropping without P fertilization (Jones, 1972; 
Adetunji, 1994), and total P as a fraction of the total soil P tends to increase with 
soil age (Walker & Syers, 1976; Cross & Schlesinger, 1995). Total P is lower in 
sandy soils common in the semiarid tropics than in medium- and fine-textured 
soils. Only a small fraction of total soil P is labile in the short term; the vast 
majority of soil P occurs in stabilized soil organic matter (SOM) and is not rapid- 
ly mineralized (Fig. 6-1). 

Use of available soil P by plants and soil biota is the driving force for the 
conversion of Pi to organic compounds and the subsequent mineralization of Po 
(Dalai, 1977; Stewart & Tiessen, 1987). Much of the P associated with soil biota 
is contained in bacteria and fungi. Amoebae, nematodes, and soil fauna general- 
ly contain only a small fraction of the P in soil biota, but they can be very impor- 
tant in the mineralization of P and the availability of P (Frossard et al, 1995). 
The Po concentration can exceed Pj concentration in soil solution due to lower 
sorption of P , and P can be the main form in which P moves in soil (Frossard et 
al., 1989). Brouwer and Powell (1995) reported that P from relatively large inputs 
of cattle manure can be leached below the rooting depth of crops on Psa 
Paleustalfs (85 to 90% sand). 


Sodium bicarbonate and sodium hydroxide are two extractants commonly 
used in attempts to isolate labile P . Bicarbonate P contains labile compounds 
that are easily mineralizable to plant-available P (Bowman & Cole, 1978). Anion- 
exchange resin has also been used to extract labile P (Rubaek & Sibbesen, 1993; 
Andersohn, 1996). NaOH P contains P associated with humic compounds and P 
sorbed to Fe and Al oxides (Cross & Schlesinger, 1995). NaOH P is more stable 
than bicarbonate P , but it can be an important P source for soil microorganisms, 
especially when labile P; is low (Chauhan et al., 1981). 

In a comparison of fallow and cropping systems on an Oxisol in Brazil, the 
bicarbonate P fraction remained relatively constant regardless of cropping histo- 
ry, whereas the NaOH P fraction reflected changes in SOM due to cultivation 
(Tiessen et al., 1992). Beck and Sanchez (1994) concluded for an nonfertilized, 
18-yr cropping system on an Ultisol in Peru that NaOH P was the dominant 
source of plant-available P. In fertilized systems, an the other hand, NaOH Pj was 
the dominant fraction related to plant availability of P. Maroko et al. (ICRAF, 
1996, unpublished data) measured NaOH P without preceding resin and bicar- 
bonate extractions, after a 1.5-yr nonfertilized sesbania [Sesbania sesban (L.) 
Merr.] tree fallow, a 1.5-yr uncultivated natural fallow, and nonfertilized maize 
(Zea mays L.) monoculture on a Kandiudalific Eutrudox with low labile P; (bicar- 
bonate-extractable Pj = 2 mg kg" 1 ) in Kenya. NaOH P , which included bicar- 
bonate Po, differed among the fallow and maize systems, whereas bicarbonate P 
and Pj were not affected by the systems. NaOH P without bicarbonate P , as 
determined by difference between the two pools, was affected by the systems (P 

In summary, NaOH P appears to represent a slowly labile P pool, which is 
an important source of mineralizable P within the time frame of 2 to 10 yr in 
tropical soils. Tiessen et al. (1994), however, cautioned that the term labile, as 
associated with specific extractants, should not be universally applied to tropical 
and temperate soils and to short and long time frames. Moreover, the P measured 
in specific extracts is not exactly congruent with P in conceptual pools illustrat- 
ed in Fig. 6-1 (Gijsman et al, 1996). 


Inorganic Phosphorus 

The supply of P from P; to plants depends on (i) the concentration of Pj in 
soil solution, (ii) the quantity of solid-phase Pj that serves as a reserve to replen- 
ish P in soil solution, and (iii) the ability of the soil to maintain the solution P con- 
centration (Holford, 1997). A portion of the Pj sorbed by Fe and Al oxides can be 
reversibly released to replenish P, in soil solution, although desorption of P; is 
slower than sorption of P; (Barrow, 1983). Desorption of P in a soil is inversely 
related to the duration of P sorption, content and form of Fe and Al oxides in the 
soil, soil P-sorption capacity, and the portion of soil-sorbing capacity that is unoc- 
cupied (Frossard et al., 1995). Some organic compounds, such as oxalate and 
malate, can desorb P through competition with P-sorption sites and complexation 


of Fe and Al ions in acid soils containing Fe and Al oxides (Lopez-Hernandez et 
al, 1986; Hue, 1991; Fox & Comerford, 1992). 

The supply of Pj to plants is usually estimated by extractive tests that mea- 
sure the Pj in solution and an amount of labile Pj. Chemical extractants, however, 
do not consider the kinetics of P t sorption and desorption, which can strongly 
influence the supply of P to plants (Kuo, 1990; Raven & Hossner, 1994). The 
inclusion of an index ofP sorption with chemically extracted Pj can improve the 
estimation of P availability (Indiati & Sharpley, 1996). The content of soil clay 
within a region of similar parent material is one such index of P sorption. The 
inclusion of clay content with extractable Pj improved the prediction of P fertil- 
izer requirements on soils in Brazil (Lins et al, 1985; Lins & Cox, 1989). 
Recovery of P fertilizer by upland rice (Oryza sativa L.) was inversely related to 
the silt plus clay content of acid soils in Cote d'lvoire (Van Reuler, 1996). This 
use of clay content appears effective on soils with predominantly kaolinitic clay, 
in which Fe and Al oxide content is directly related to the clay content. It will not 
be effective on soils with free calcium carbonate or P sorption by amorphous alu- 
mino-silicates and humus-Al complexes (Cox, 1994). 

The critical concentration of extractable P required for a given quantity of 
P uptake by a plant decreases with increasing P sorption. Cox (1994), for exam- 
ple, showed that the critical P level with Mehlich-3 extractable P was 22 mg P 
L J for soil with low P sorption (10% clay) and 13 mg P L'" for a soil with high- 
er P sorption (25% clay). 

Phosphorus sinks, such as resins and Fe-oxide impregnated paper (Sharpley 
et al, 1994; Menon & Chien, 1995; Menon et al., 1997), and isotope exchange 
(Fardeau et al., 1996; Morel et al., 1996; Di et al, 1997) generally estimate P 
availability to plant roots more closely than do chemical extractants. Lins and 
Cox (1989), however, showed that the inclusion of an index of P sorption (clay 
content) with even resin-extractable P increased the predictability of P supply. 

Organic Phosphorus 

Net mineralization of P is generally directly related to total soil P for both 
fertilized and nonfertilized soils (Sharpley, 1985). Net mineralization of P j there- 
fore, tends to be more important as a source of plant-available P on highly than 
slightly weathered soils because of the generally greater total P in highly weath- 
ered soils (Tiessen et al., 1984; Sharpley etal., 1987). 

Early research on soils from Kenya (Friend & Birch, 1960) and southern 
Nigeria (Adepetu & Corey, 1976) revealed a strong direct relationship between 
total soil P„ and plant-available P. The mineralization of P can raise Pj in soil 
solution, but the Pj can be rapidly sorbed on Fe and Al oxides (Adepetu & Corey, 
1977). Declines in total SOM, as traditional rotational systems containing fallows 
are replaced by continuous cropping, can lead to reduced supply of plant-avail- 
able P from mineralization of Po and thus greater need for external inputs of P. 

The cycling of Po to plant-available P is a function of the size and activity 
of the microbial biomass (Stewart & Tiessen, 1987), which in turn are governed 
by the supply of available C and nutrients. In many ecosystems including small- 
holder farms in Africa, C is the factor most limiting microbial biomass and rapid 


r (Smith, 1994). The cycling of P through plants and soil biota 
can consequently be closely linked with C cycling (Huffman et al., 1996). In 
highly P-deficient soils with a wide range of C-to-P ratios in organic pools, how- 
ever, the relationship between P and C cycling may be less closely linked 
(Gijsman et al., 1996). 

Possible indicators of P dynamics and conversion of P to P; include 
microbial biomass, biologically active SOM, and inputs of mineralizable organic 
material (Tiessen et al., 1994). Maroko et al. (ICRAF, 1996, unpublished data) 
found that microbial biomass P and P associated with light-fraction SOM differed 
among an nonfertilized tree fallow, an uncultivated natural fallow, and continu- 
ous maize after 1.5 yr on an Eutrudox with low labile Pj. Microbial biomass P and 
light-fraction P also were each correlated with yield of subsequent nonfertilized 
maize (ICRAF, 1996, p. 69-70; ICRAF, 1996, unpublished data). 

Bowman and Cole (1978) proposed that the sum of bicarbonate Pi and P 
would be a better indicator of plant response to P than bicarbonate P, alone, espe- 
cially for soils low or deficient in P based on bicarbonate Pj. Soils with low bicar- 
bonate P; and much higher bicarbonate P , however, can be responsive to P fer- 
tilizer. An Eutrudox in western Kenya with 1 to 3 mg bicarbonate P, kg" 1 and 1 1 
to 15 mg bicarbonate P kg" 1 in the top 15-cm soil layer, for example, was high- 
ly responsive to P (J. Maroko, ICRAF, 1996, unpublished data). In contrast to 
other findings on acid soil without P fertilization (Linquist et al, 1997a), bicar- 
bonate P was not related to maize yield. This observation suggests that a portion 
of the bicarbonate P was not readily mineralizable. 

The importance of P„ as a source of plant-available P might explain the 
poor correlations between crop response to applied P and P extracted by conven- 
tional soil P tests that do not assess P (Warren, 1992; Tiessen et al., 1994). Such 
poor correlations could arise when the mineralization of P supplies sufficient 
plant-available P to match or exceed the shortfall in P release from labile Pi to 
meet plant demand for P. When labile Pi supplies sufficient P to meet plant 
demand, the combination of an indicator of P availability from P with 
extractable Pi did not improve the correlation with plant growth (Mnkeni et al., 
1995). Extractable Pi together with a labile P -to-labile Pj ratio, might be valuable 
in assessing the importance of P mineralization. The relative importance of P 
mineralization, compared with desorption of Pj, as a source of plant-available P 
would presumably be greatest at low extractable P i; high labile Po-to-labile Pj 
ratio, and high microbial P. 


Plants differ greatly in their ability to grow on soils with low P status and 
to respond to P inputs (Sanchez & Salinas, 1981). These differences are related 
to the efficiency of plants to take up and use soil P. Factors affecting uptake of P 
include the abilities of plants (i) to absorb P from low soil solution concentra- 
tions, (ii) to explore a large soil volume, (iii) to solubilize soil Pi through pH 
changes and the release of chelating agents, and (iv) to release phosphatase 
enzymes (Kirk et al, 1993; Lajtha & Harrison, 1995). 


The solution P concentration required for optimum growth varies among 
plants (Sanchez & Uehara, 1980). A value of 0.2 mg P L 1 has been used as a 
standard solution concentration for comparing P sorption of different soils (Juo 
& Fox, 1977), but it should not be viewed as applicable to all crops. Moreover, 
its use as an indicator of P requirements can result in higher P applications than 
required for maximum crop yield. The uptake of soil P can be enhanced through 
association of plant roots with mycorrhizas. The extensive proliferation ofmyc- 
orrhizal hyphae enables the exploration of a greater soil volume (Lajtha & 
Harrison, 1995). 

The availability of soil P can be increased by organic acids and acid phos- 
phatases exuded by plant roots (Delhaize, 1995; Randall, 1995) and by plant- 
induced changes in soil pH (Marschner, 1995). Organic acids in root exudates can 
complex Fe and Al, resulting in release of P bound by Fe and Al oxides (Ae et 
al., 1990; Otani etal., 1996). Hedley etal. (1994) reported that most soil P taken 
up by rice as a result of root-induced changes in soil pH and release of P-solubi- 
lizing agents was from NaOH P ; . Gahoonia and Nielsen (1992) found root- 
induced depletion of both NaOH P ; and P by rape (Brassica napus L.), suggest- 
ing that short-term mineralization of P in the rhizosphere supplied P to the plant. 
The secretion of H + and acidification of the rhizosphere are greater for legumes, 
which accumulate N through symbiotic N? fixation rather than through uptake of 
nitrate (De Swart & Van Diest, 1987). 


Sanchez and Palm (1996) tentatively defined soil P capital as the stock of 
soil P that gradually supplies plant-available P for about 5 to 10 yr. This stock was 
referred to as reserve capital P by Baanante (1998), but authors in this publica- 
tion (Izac, 1997, this publication; Sanchez et al., 1997, this publication) simply 
refer to it as capital P. Sanchez et al. (1997, this publication) highlighted that the 
P fluxes from capital P to P taken up by a crop during a growing season are syn- 
onymous to the concept of service flows in economics. 

The element of time is important in the definition of P capital because there 
is a component of inert P that does not become available to plants within a rea- 
sonable time frame, which for the purposes of this chapter is defined as 10 yr. 
Efforts to improve capital P also will contribute to the inert P pool. The goal of 
replenishing soil P is, therefore, not to maximize soil P stocks. Instead, the goal 
is to increase capital P to a size that results in service flows sufficient for crop 
production without P limitations for several seasons (Sanchez et al., 1997, this 

The P available to crops within one season was referred to as liquid capital 
P by Baanante (1998), but it is not capital P as defined by Sanchez and Palm 
(1996) and Sanchez et al. (1997, this publication). In this chapter, liquid P will be 
used to refer to soil P available to crops within one growing season. Liquid P 
includes service flows, the flux of P from capital P, as well as a portion of the P 
added in the cropping season as mineral fertilizers (hereafter referred to as inor- 
ganic fertilizers) and organic materials. When inorganic P fertilizers and organic 


materials are not added in the cropping season, capital P supplies the P taken up 
by crops. In that case, service flows and liquid P are synonymous. 

Capital P includes sorbed P, a portion of the primary P minerals, and a por- 
tion of the stable SOM-associated P (Fig. 6-1). Although no extractant will pre- 
cisely quantify capital P, NaOH Pi and P as determined by sequential extraction 
presumably contain P that gradually becomes available for plant use over sever- 
al cropping seasons. Liquid P is presumably approximated by labile P; and P , 
microbial P, and soil solution P. Microbial biomass P and the sum of sequential 
extractions with resin and bicarbonate presumably approximate a pool that is 
readily available for plant uptake; however, only the portion of bicarbonate P 
that is mineralized within one season would represent liquid P. 

A portion of PR-P added to soil, depending upon soil properties and the 
reactivity of the PR, becomes available for plant uptake over a number of crop- 
ping seasons. The undissolved PR that gradually releases P for up to 10 yr to liq- 
uid P represents capital P not detected by NaOH extraction. The P dissolved from 
PR like P from any water-soluble P source, is subject to conversion to plant- 
unavailable P forms due to reactions with soil components such as Al and Fe 
oxides (Fig. 6 — 1). 

The contribution of capital P to crop production depends not only on its size 
but also on the rates and magnitudes of P flows among soil P pools and on the 
ability of plants to take up and efficiently use soil P. These flows might also be 
influenced by soil aggregation. Linquist et al. (1997b), for example, reported that 
an increase in size of soil aggregates can increase availability of applied P to 
plants through reduction in sorptive surface area. 

Organic-Based Systems and Agroforestry 

Relatively large additions of high-quality organic materials can increase 
liquid P and capital P. Addition of 5.5 1 dry matter ha" 1 of tithonia [Tithonia diver- 
sifolia (Hemsley) A. Gray] biomass, equivalent to 15 kg P ha" 1 , to a Kandiudalf 
in Kenya increased resin P and NaOH P; at 2 to 16 wk after application 
(Nziguheba et al., 1998). This observation is consistent with earlier reports that 
tithonia was an effective source of nutrients, including P (Nagarajah & Nizar, 

Low-quality organic materials (see Palm et al, 1997, this publication), 
however, might be ineffective sources of plant-available P. Application of maize 
stover equivalent to 15 kg P ha" 1 , for example, had little effect on resin P and no 
effect on NaOH Pj in 16 wk, apparently because maize stover has less soluble Pj 
and slower decomposition than tithonia (Nziguheba et al., 1998). 

Even though some organic materials have the potential to increase liquid P 
and capital P, the amount of P that can be added through organic materials is 
restricted by the limited supply of organic materials at the farm level (Palm et al., 
1997, this publication). For example, application of 18 kg P ha- 1 , the approximate 
amount of P used by a 2 t ha" 1 maize crop (Palm, 1995), would require applica- 


tion of 6 t dry matter ha" of organic material containing 3 g P kg" . Low avail- 
able soil P, however, can limit the production of plant biomass for use as organic 
inputs. Organic materials also frequently have a higher N-to-P ratio than the ratio 
required by crops (Palm, 1995). Jama et al. (1997) found in western Kenya, that 
rather than using organic material with a high N-to-P ratio to supply all the P 
required by maize, it was economically more attractive to integrate TSP with the 
organic material, whereby the organic material provided the required N for the 
crop and the TSP met the additional requirement for P. Integrated application of 
inorganic P fertilizer with organic material would have an agronomic advantage 
compared with sole application of inorganic P, if the organic material enhanced 
the availability of added P (Palm et al., 1997, this publication). 

Maize monoculture 

1 66 || 3.8 II 23 272 
| NaOH Pi | Bicarbonate P ( | Bicarbonate P„ NaOH P„ 

1 10 - 7 II 1 " 2 I 
| Microbial P 1 1 Light fraction P | 

P removed = 14.9 NaOH + bicarb P = 364 
P recycled = 

Sesbania fallow 

















63 | 
NaOH P, | 

3.7 21 296 
Bicarbonate P, | Bicarbonate P | | NaOH P | 

17.7 II 3.6 | 
Microbial P 1 1 Light fraction P | 

P removed = 11.9 
P recycled = 6.7 

NaOH + bicarb P = 383 

Fig. 6-2. Effect of ma 
2 yr. All values are 

export and cycling of plan! P for 2 yr (4 seasi 
n kilograms or P per hectare 


tion of 6 t dry matter ha"' of organic material containing 3 g P kg" . Low avail- 
able soil P, however, can limit the production of plant biomass for use as organic 
inputs. Organic materials also frequently have a higher N-to-P ratio than the ratio 
required by crops (Palm, 1995). Jama et al. (1997) found in western Kenya, that 
rather than using organic material with a high N-to-P ratio to supply all the P 
required by maize, it was economically more attractive to integrate TSP with the 
organic material, whereby the organic material provided the required N for the 
crop and the TSP met the additional requirement for P. Integrated application of 
inorganic P fertilizer with organic material would have an agronomic advantage 
compared with sole application of inorganic P, if the organic material enhanced 
me availability of added P (Palm et al., 1997, this publication). 

Maize monoculture 

Soil (0 to 15 cm) 

I 66 II 3.8 M 23 || 272 I 
| NaOH P, 1 1 Bicarbonate P, 1 1 Bicarbonate P„ 1 1 NaOH P | 

Microbial P 

1.2 | 
Light fraction P | 

P removed = 14.9 
P recycled = 

NaOH + bicarb P = 364 

Sesbania fallow 


















63 | 
NaOH Pi 1 

3.7 21 296 
Bicarbonate P, [ [ Bicarbonate P [ | NaOH P 

17.7 || 3.6 
Microbial P 1 1 Light fraction P 

P removed = 11.9 
P recycled = 6.7 

?ig. 6-2. Effect of maize monoculture and 
in western Kenya on export and 
2 yr. All values are in kilogram; 

NaOH + bicarb P = 3 

e-sesbania fallow systi 


The recovery of added TSP-P at 10 mo (Fig. 6-3) was 11% in bicarbonate 
Pi and 59% in NaOH Pj. The latter is similar to the 52 to 58% recovery of added 
TSP-P in NaOH P, on an Ultisol with 60% clay (Linquist et al., 1997a). On soils 
with lower P-sorption capacity, the recovery of fertilizer P tends to decrease in 
the NaOH Pj fraction and increase in more labile fractions (resin and bicarbonate 
Pj). The NaOH Pj, nonetheless, is a primary sink for fertilizer P even on low P- 
fixing soils (Aulakh & Pasricha, 1991). 

Small applications of soluble P fertilizer to P-deficient soil with low or 
moderate P-sorption capacity can increase crop yields with little or no detectable 
build up in extractable soil Pj. Broadcast application of 25 kg TSP-P ha' 1 and 
placement of 10 kg TSP-P ha" 1 in the planting hole, for example, increased maize 
yield in the two cropping seasons after P application on an Eutrudox with 42% 
clay in western Kenya (Table 6-1). These low P rates had no detectable effect on 
liquid P in the bulk soil, as estimated by resin P. A build up of liquid P and cap- 
ital P, however, may have occurred in the vicinity of fertilizer granules. 

A larger application ofTSP (>50 kg P ha" 1 ) increased resin P in the bulk 
soil at the end of the first maize cropping season (5 mo after P application; Table 
6-1). Resin P rapidly decreased between the end of the first and second seasons 
after P application, and the percentage decrease was greater for high (100-250 kg 
ha 1 ) than moderate (50 kg ha" 1 ) P rates. Other researchers have similarly found 
greater percentage declines in bicarbonate Pj and double acid extractable P (Cox 
et al, 1981) and Mehlich-1 extractable P (Linquist et al, 1996) following high 
than low rates of P application. 

Application of PRs classified as medium or highly reactive (Diamond, 
1979; Hammond & Leon, 1983) can increase capital P, as determined by NaOH 
Pj, in acid soils. For PRs, reactivity is indicative of the rate and extent of PR dis- 
solution in soil and is determined by their inherent mineralogy. Application of 


Table 6-1. Effect of rate and method of application of triple superphosphate (TSP) on s 
maize grain yield, and added net benefit on an Eutrudox in western Kenya (Palm e 
unpublished data). 

Added nel: benefitt 

2 25 Broadcast 4 4 4.4 1.5 219 212 

3 50 Broadcast 7 6 4.4 1.7 148 173 

4 100 Broadcast 17 6 5.6 2.0 216 284 

5 150 Broadcast 20 9 5.4 1.4 34 17 

6 250 Broadcast 33 16 5.1 1.7 -286 -259 

7 10 Seed placed§ 4 4 3.9 1.6 242 243 
SED1 2.7 2.5 0.85 0.42 143 181 

Phosphorus was applied only in the first season. 

Urea was applied at 100 kg N ha 1 to all plots in the first season. The added net benefits were deter- 
in n i I \ i rtial budg in in hich on] it n benefit i i iri I from the t ontrol with no 
added P and no added urea (i.e., costs of fertility-enhancing inputs and the value of increased maize 

yield) were consult , d. Grain yield 1 th \va li in 'n n I ason and 1.5 t ha ' 

nil) hi i i I li in I i i i i I in i I. ( 1997) 

Phosphorus fertilizer was mixed with soil in the planting hole at the time of planting. 
SED, standard en oi ftli (Jiff m ii in n i oi lf= 12. 

250 kg P ha" 1 as reactive Minjingu PR increased bicarbonate plus NaOH Pj by 17 
mg P kg" 1 at 2 wk after application on an acid soil in Kenya. The recovery of 
added P in bicarbonate plus NaOH Pj steadily increased to 50 mg P kg"' at 10 mo 
(Fig. 6-3). The recovery of added PR-P in NaOH P, was 8% at 2 wk and 27% at 
10 mo. 

The P from water-soluble P sources (e.g., TSP) is rapidly released to labile 
soil P and then to capital P, as determined by NaOH P t . The P in most PRs is only 
slowly released in acid soils resulting in a gradual build up of NaOH Pi(Fig. 6-3). 
Substantial yield increases observed in the year of application of reactive PRs 
indicate that reactive PRs also increase liquid P and supply P to the crop in the 
first season (Bationo et al., 1986; Heliums et al., 1992). The NaOH P„included 
in Fig. 6-3, underestimates total capital P following PR application because it 
does not include undissolved PR, roughly estimated with HC1 P, that gradually 
releases P for a number of seasons. 

Soil Phosphorus Requirements 

Based on P sorption isotherms for 200 soils from West, East, and southern 
Africa, Warren (1992) concluded that fertilizer P requirements tend to follow the 
order Andisols > Oxisols > Ultisols > Alfisols > Entisols. There is much vari- 
ability in P requirements among soil orders, but for soils other than Andisols there 
is generally a direct relationship between P requirements and surface area of Fe 
and Al oxides, which is a function of clay content and mineralogy. Soils derived 
from volcanic ash (Andisols) characteristically have high P sorption, resulting in 
low recovery of added P (Sanchez & Uehara, 1980; Vander Zaag & Kagenzi, 


1986). Because of the irreversible nature of P sorption in these soils, there is typ- 
ically little residual effect of P applications. Andisols are important in localized 
areas in East and Central Africa, but they are relatively unimportant overall in 
Africa. Some soils near young volcanos, on the other hand, have received vol- 
canic ash. Such soils with volcanic admixtures can be relatively fertile and high 
in available P (Wielemaker & Boxem, 1982). 

Large areas of Africa, particularly in the semiarid tropics, are dominated by 
sandy soils with low P sorption and hence low fertilizer P requirements 
(Mokwunye et al., 1986; Warren, 1992). A modest annual application of 15 to 20 
kg P ha" is usually adequate for these soils (Bationo & Mokwunye, 1991). 
Baidu-Forson and Bationo (1992), in an evaluation of a 6-yr experiment on a 
Psammentic Paleustalf (94% sand) in Niger, recommended annual application 
rates of 17.5 kg P ha ' as SSP in combination with N and K fertilizer. When 5 t 
manure ha was applied every third year, the annual P application could be 
reduced to 8.7 kg P ha ' as SSP. 

Warren (1992) found more varied P sorption among soils from East and 
southern Africa than from West Africa, but many soils in East Africa have medi- 
um to high P requirements, based on P sorption isotherms. Phosphorus sorption 
is not routinely determined in African laboratories or elsewhere, but P-sorption 
capacity of acid soils can be generally approximated from Fe and Al oxide con- 
tent, clay content, and red color. Braun et al. (1997), therefore, used relatively 
easily obtainable information on soil clay content and color to develop indicators 
of potential P fixation for soils in the highlands of East Africa based on the 
Fertility Capability Classification (Sanchez et al., 1982). They used the criteria 
that highly P-fixing soils were Andisols (Andosols in FAO/UNESCO, 1977) and 
moderately P-fixing soils had > 35% clay and Munsell 5YR or redder. Based on 
available data for 63% of the highlands (1200-3300 m altitude and > 400 mm 
rainfall for 5 consecutive months) in Kenya, about 50% of the soils in the high- 
lands were moderately P fixing and 16% were highly P fixing. In Ethiopia, 36% 
of the highland soils were moderately P fixing and 1% was highly P fixing. 

Based on soil nutrient depletion (Smaling et al., 1997, this publication) and 
P-sorption capacity, the highly weathered soils in the densely populated and 
intensively cultivated highlands of East Africa typically have higher fertilizer P 
requirements than the low P-fixing soils in West and southern Africa. In subse- 
quent discussions in this chapter, reference to P-sorption capacity will be based 
on P sorption isotherms (Fox & Kamprath, 1970) where sorbed P at 0.2 mg solu- 
tion P L -' is <100 mg P kg-' soil for low P-fixing soil, 100 to 400 mg P kg ' soil 
for moderate P-fixing soil, and >400 mg P kg" ' soil for high P-fixing soil. Some 
soils referred to as P fixing by Sanchez et al. (1997, this publication; see Fig. 1-3) 
are moderate P fixing by our definition. 

Relatively small rates ofP fertilizer (10-25 kg P ha '), particularly when 
mixed with soil in the planting hole, can increase crop yield and be financially 
attractive on moderate P-fixing soils in the highlands of East Africa (Table 6-1 ; 
Jama et al., 1997). Crop yields can be further increased with higher P rates, which 
can also be financially attractive (e.g., 100 kg P ha" 1 , Table 6-1). Measurements 
of crop response and financial benefits for P fertilizer in farmers' fields can be 
highly variable, as illustrated by the large standard error of the difference in 


means (SED) in Table 6 - 1 . The variability arises from heterogeneity not only in 
soil fertility but also in other crop-limiting factors, such as diseases and pests 
(e.g., incidence ofStriga sp.). 


The sources of P that can be provided in sufficient quantities for building 
soil P capital are essentially limited to soluble P fertilizers and some PRs that are 
capable of supplying P on acid soils within a reasonable time frame of 10 yr. As 
indicated previously, organic materials are typically too low in P content and too 
limited in supply to represent a substantial source of P for building soil P capital. 
Organic materials are more suited as supplements with inorganic P sources (see 
Palm et al., 1997, this publication). 

Indigenous Phosphorus Sources 

In some cases the direct application of indigenous PR as a source of capi- 
tal P is viewed as an attractive option for building soil P fertility because it poten- 
tially involves lower production costs and capital investments than the production 
of water-soluble P fertilizers from indigenous PR sources (Hammond et al., 
1986b; Rajan et al., 1996). Phosphate rock deposits are scattered throughout 
Africa (Fig. 6 — 4). Deposits in West Africa are predominantly sedimentary, where- 
as those in East Africa are predominantly of igneous origin. Exceptions in East 
Africa include Minjingu PR a sedimentary deposit in Tanzania, and lies Barren 
PR a relatively small guano deposit in Madagascar. 

Generally speaking, sedimentary ores are more soluble and tend to be less 
costly to excavate and process than igneous ores. Some agromineral resource 
evaluations and site-specific phosphate deposit studies have been conducted in 
Africa (Savage, 1987; Van Kauwenbergh, 1991), but verified details on the size 
and composition for many of the deposits are fragmented (Table 6-2). In such 
cases, the limited availability of geologic information has prevented an econom- 
ic analysis on the use of the PRs. There is a need to update the information on 
PRs presented in Fig. 6-4 and Table 6-2. 

Phosphate rocks with low reactivity (Table 6-2) will require additional pro- 
cessing of the PR raw material before the material can be used in improving soil 
P capital within a reasonable time frame of 10 yr. Partial acidulation of PR is an 
alternative use for indigenous PR that is too low in reactivity for direct applica- 
tion (Bolan et al., 1990; Rajan & Marwaha, 1993). The production of partially 
acidulated PR (PAPR) uses only a portion of the acid required for production of 
SSP and hence has a lower cost than SSP (Schultz, 1986). PAPR has been shown 
to be effective on sandy soils with low P sorption in West Africa (Chien & 
Menon, 1995a). The water-soluble P from PAPR presumably promotes early root 
growth following application of PAPR thereby enabling more effective use of P 
from PR (Hammond et al., 1986b). The factory-gate cost of P from sulfuric acid- 
based 50% PAPR is about 80% of the cost of P from SSP (Schultz, 1986). 
Phosphorus content is lower for PAPR than for high-analysis fertilizers such as 



1 Cabinda 


20 Sekondi 


40 KalkteW 

2 Longonjo 



21 Bambuta-Bomi Hill 

41 Offshore 




4 Mekrou 

22 lies Barren 

42 HaJaib 



43 Sherelk 

Burkina Fa so 

23 Chilwa Island 


6 Arty 


44 Minjingu 

7 Diapega-Kodjari 


45 Panda Hill 


25 Assakerei 


46 Akoumape-Hahotoe 


26 Tllemsi 





27 Bofal-Loubboira 

47 Busumbu 

Cen . Africa Rap. 


48 Sukulu 


28 Maputo 

Zaire (now Congo 
Dem. Hep.) 

49 Bingo 

Congo Republic 

29 Muambe 

30 Aschia Tlnamou 


50 Lueshe 


31 Tahoua 


1 4 Bikilal 



51 Rufunsa 

52 Nkombwa 




16 Iguela 


34 Gambia Name! 

54 Shawa 

Guinea Bissau 


17 Saliquinba 

36 Taiba-Tbies 


37 Ziguinchcw 


38 El Bur 

Table 6-2. Estimated 

of some phosphate rocks (PR) in sub-Saharan Africa. 

Burkina Faso 
Congo D.R. 




South Africa 


Cabinda and Lucunga 



Tahoua and Pare W 
Taiba and Thies 


['a, Ida Hill 
Sukulu Hill 


















t Sources, McClellan and Notholt (1986). Sheldon 1 1 987). and Van Kauwenbergh (1995) unless ir 
Reacth ity .it ,i PR is determined by its inhere nt niineralogx anil is correlated to the amount of citra 
um reactivity = 15 to 25 g CS P kg" 1 , and low reactivity <15 g CS P kg" 1 . 

§ From FAO (1996), except for Minjingu PR. 


§ NA, not available, 
tt Extraction in 1995. Production halted in 1996-1997 

te soluble (CS) P in the PR: High reactivity >25 g CS P kg" 1 , medi- 

closure ofTanga fertilizer factory (E.P. Mwaluko, Mipco Ltd, Nairobi, 1997, f 


TSP, diammonium phosphate (DAP), and monoammonium phosphate (MAP). 
Therefore, transport cost will be relatively higher for PAPR, and adversely affect 
its economic attractiveness. 

Partial acidulation is not suitable for PRs with high Fe and Al oxide con- 
tents. The Fe and Al oxides react with water-soluble P formed during the acidu- 
lation of PR to form water-insoluble forms of P (Hammond et al., 1989; Chien & 
Menon, 1995a). The removal of Fe and Al oxides from raw PR ore by beneficia- 
tion can result in a high-grade ore suitable for partial acidulation (Butegwa et al., 
1996a), but this substantially increases the cost (IFDC, 1990, unpublished data). 

An alternative to partial acidulation of low-reactive PR is compaction of the 
PR with water-soluble P fertilizer, such as TSP (Lupin & Le, 1983; Menon & 
Chien, 1996). Compaction with soluble P fertilizer and partial acidulation are 
equally effective for low-reactive PR with low Fe and Al oxides; but compaction 
with soluble P is more effective than partial acidulation for PR that is high in Fe 
and Al oxides (Menon & Chien, 1990; Menon et al., 1991, 1995). The relative 
effectiveness of PAPR and compacted PR plus soluble P fertilizer, with respect to 
soluble P fertilizer alone, has been reported to be greater on soils with high than 
low P-sorption capacity (Mokwunye & Chien, 1980; Chien & Hammond, 1989). 
Butegwa et al. (1996b), however, recently reported that PAPR and compacted PR 
plus TSP prepared with unreactive Sukulu Hills PR from Uganda did not increase 
in relative effectiveness with increasing soil P-sorption capacity on soil limed to 
P H6.0. 

The physical combination of water-soluble P with PR, as in PAPR and 
compacted PR plus soluble P fertilizer, can result in chemical interactions that 
increase the effectiveness of P use from PR by crops (Mokwunye & Chien, 
1980). When both water-soluble and water-insoluble P are present, some acid 
produced from dissolution of the soluble P source is neutralized by the PR and 
hence does not react with Fe and Al minerals in the soil to form insoluble Fe-Al 
phosphates. The reaction of acid with PR might also release additional P into 
solution (Mokwunye & Chien, 1980; Chien & Menon, 1995a). Recent results 
from a pot study by Chien et al. (1996) indicate that soluble P fertilizer can 
enhance the effectiveness of PR even when the soluble P and PR are not physi- 
cally combined. Phosphorus uptake from PR by maize and cowpea [Vigna 
unguiculata (L.) Walp. sp. unguiculata] in a greenhouse study was increased by 
a physically separate application of TSP. The water-soluble P from TSP presum- 
ably enhanced early root development, which enabled more effective utilization 
of P from PR than when PR was applied alone. These pot study results require 
q and verification under field conditions. 

Commercial Phosphorus Sources 

Phosphorus fertilizer consumption in Africa is relatively low on an area 
basis and largely concentrated in a few countries. Total P fertilizer consumption 
in 1994 — 1995 in sub-Saharan Africa, excluding South Africa, was 0.17 x 10 6 
tonnes P (FAO, 1996). This represents only 38% of the total P consumption in 
continental Africa, and it is only marginally greater than the P consumption in 
South Africa alone (0.13 x 10 6 tonnes P). Nigeria, Kenya, Zimbabwe, and 


Ethiopia accounted for 66% of the P consumption in 1994-1995 in sub-Saharan 
Africa, excluding South Africa. 

Phosphorus imports in 1994-1995 in sub-Saharan Africa, excluding South 
Africa, (0.17 x 10 6 tonnes P) were comparable to P fertilizer consumption. Some 
of the imported P, particularly in Nigeria, was used in manufacturing fertilizers 
(Gerner & Harris, 1993). The most important sources of P fertilizer in Africa are 
SSP, TSP, ammonium phosphates, and other compound fertilizers. These soluble 
P fertilizers are normally comparable in agronomic effectiveness per unit of P. 
High-analysis fertilizers, however, are favored because of their relatively lower 
transportation cost per unit of P, although SSP can be a desirable fertilizer when 
sulfur is required. 

Even though PR deposits are found throughout sub-Saharan Africa, only 
those in Senegal, South Africa, Togo, and Zimbabwe have been developed on a 
large commercial scale (Table 6-2). In Togo, the PR is exported. In Senegal and 
South Africa, the PR is both exported and used in the domestic manufacture of 
fertilizers. In Zimbabwe, the PR is used solely for the domestic manufacture of 
SSP and TSP. The SSP is marketed directly, whereas the TSP is either used in 
compound fertilizers or mixed with SSP to form double superphosphate (Johnsen 
etal, 1996). 

Targeting Inorganic Phosphorus Sources 

The suitability of a PR for direct application to soil depends upon the min- 
eralogy and reactivity of the PR, soil properties, the crop, and the economics of 
use associated with the PR. Only PRs with high or medium reactivity are poten- 
tially suitable for direct application. The PR must be incorporated into the soil; 
and except in the case of highly reactive PRs, the PR must be finely ground. High 
soil solution Ca slows down the dissolution of PR because Ca ions are released 
during PR dissolution (Robinson & Syers, 1990). The effective dissolution of PR 
in soil requires low soil pH, low soil exchangeable Ca, and low soil solution P 
concentration (Khasawneh & Doll, 1978; Chien & Menon, 1995b; Rajan et al., 
1996). Direct application of PR is normally not recommended for low rainfall 
areas, due to erratic agronomic effectiveness under conditions of low soil water 
content (Hammond et al, 1986b). 

The dissolution of PR increases as soil P-sorption capacity increases 
(Smyth & Sanchez, 1982; Mackay et al., 1996), but extractable P also decreases 
with increasing soil P-sorption capacity (Syers & Mackay, 1986; Kanabo & 
Gilkes, 1987). The short-term agronomic effectiveness of both PR and soluble P 
fertilizer decreases with increasing P-sorption capacity due to decreasing soil 
solution P (Mokwunye & Chien, 1980). The determination of relative effective- 
ness of PR and soluble P fertilizer on soils with varying P sorption, however, has 
been confounded by differences in other soil properties, such as pH and 
exchangeable Ca (Chien et al., 1980; Hammond et al. 1986b). Hammond et al. 
(1986a), in a greenhouse experiment, altered the P sorption of a soil through the 


addition of amorphous Fe gel while holding other soil properties constant. They 
found that the effectiveness of PR relative to TSP decreased with increasing P- 
sorption capacity. Although a high P-sorption capacity can promote more rapid 
dissolution of PR, the low soil solution P concentration resulting from high P 
sorption may limit plant growth (Mokwunye & Hammond, 1992). 

The effectiveness of plant use of P from PR varies among cultivars 
(Ankomah etal., 1995) and species (Flach et al, 1987; Bekele & Hefner, 1993). 
Plants can enhance the dissolution of PR through acidification of the rhizosphere 
(Hinsinger & Gilkes, 1996), high uptake of Ca (Hinsinger & Gilkes, 1997), secre- 
tion of organic acids that complex Ca (Hoffland, 1992), and depletion of P in soil 
solution (Rajan et al., 1996). The relative effectiveness of PR compared with sol- 
uble P fertilizers also is generally greater for long-duration crops and perennials 
man for short-duration crops (Pushparajah et al., 1990; Sale & Mokwunye, 1993). 

A few PR deposits, such as the Tilemsi in Mali (Henao & Baanante, 1997), 
Tahoua in Niger (Bationo & Mokwunye, 1991), and Minjingu in Tanzania 
(Anderson, 1970; IFDC, 1990), are sufficiently reactive for direct application. 
Many other deposits, such as Pare W in Niger (Bationo et al., 1990), Hahotoe in 
Togo (Kpomblekou et al., 1991), and Sukulu Hills in Uganda (Butegwa et al., 
1996a) are not sufficiently reactive for direct application. As indicated previous- 
ly, in order for low-reactive PRs to be used as a P source, additional processing 
is required; however, additional processing (e.g., partial acidulation or com- 
paction) increases cost significantly and results in a product where the soluble P 
content as a percentage of total P remains considerably lower than for high-analy- 
sis, soluble P fertilizers. 

Application Strategies 

The agronomic effectiveness of PR is frequently determined from the com- 
parison of PR and a soluble P fertilizer at relatively large, one-time applications. 
Large applications of soluble P can have long-term residual effects on crop yield 
(Kamprath, 1967; Sanchez & Salinas, 1981), liquid P, and capital P (Linquist et 
al, 1997a). Crop yields (Janssen et al., 1987; Wolf et al., 1987) and extractable P 
(Linquist et al, 1996), nonetheless, eventually decline following large applica- 
tions of soluble P, and the one-time application of PR may have a longer term 
residual value than an equivalent one-time application of soluble P fertilizer 
(Rajan etal., 1996). 

Predicted crop yield, relative to yield when P is not limiting, for about a 10- 
yr time frame is illustrated conceptually in Fig. 6-5a for one-time applications of 
soluble P fertilizer and contrasting PRs. Relative yield will be somewhat less for 
high-reactive PR than for soluble P in the years immediately following applica- 
tion, but yield may eventually be comparable and slightly greater with the reac- 
tive PR. Relative yield will be considerably less for low-reactive PR than for sol- 
uble P following application, but with time yield may be greater with PR; how- 
ever, cumulative yield would be considerably less for the low-reactive PR than 
for soluble P. The relationships illustrated in Fig. 6-5a will depend upon soil 
properties, cropping system, and climatic factors (Rajan et al., 1996). The resid- 
ual benefit from P fertilizers, for example, would increase with increasing P sorp- 


tion. Therefore, the anticipated decline in relative yield following one-time P 
applications would be less rapid as P-sorption capacity increases. 

A study in Niger confirmed that medium-reactive PR can be nearly as 
effective as soluble P fertilizer in the seasons immediately following application, 
whereas low-reactive PR was much inferior to soluble P fertilizer. The relative 
agronomic effectiveness (RAE), defined as (yield increase with PR/yield increase 
with soluble P) x 100, following a one-time application of medium-reactive 
Tahoua PR and SSP ranged from 82% in Year 1 to 104% in Year 3 with pearl mil- 
let [Pennisetum glaucum (L.) R. Br.]. The RAE for low-reactive Pare W PR was 
47% in Year 1 and 31% in Year 3 (Bationo & Mokwunye, 1991). These results 
indicate that medium- to high-reactive PRs rapidly contribute P to both liquid P 
and capital P, as shown for Minjingu PR in Fig. 6-3. 

Bromfield et al. (1981), in a comparison of a one-time application of medi- 
um- to high-reactive Minjingu PR and SSP to maize on an acid soil in western 
Kenya, reported a RAE for PR of only 26% in the first season. When examined 
over five successive seasons, however, the cumulative RAE for PR was 75%. The 
low RAE of Minjingu PR reported after one season (26%) by Bromfield et al. 
(1981), however, may be a result of band rather than broadcast application of the 
PR. Banding of PR is frequently less effective than broadcasting and incorporat- 
ing PR (Chien & Menon, 1995b). 

ii adapted from Rajan et al.. 1996). 

n moderate P-fixing sc 


Seasonal applications of soluble P fertilizers might be more attractive than 
one-time applications for building capital P on low to moderate P-fixing soil 
(Bationo & Mokwunye, 1991). Seasonal applications of relatively small amounts 
of soluble P would enable the use of P on a larger land area at rates on the steep- 
er part of the P response curve. Predicted crop yield from seasonal application of 
soluble P fertilizer, relative to yield when P is not limiting, will remain constant 
or gradually increase with time (Fig. 6-5b), depending on the rate of P applica- 
tion. Cumulative yield could eventually be greater for seasonal than for a one- 
time application of soluble P fertilizer. On high P-fixing soils, however, relative- 
ly small applications of soluble P can be ineffective in increasing yield (Sanchez 
& Salinas, 1981). 

Crop yield with low- to medium-reactive PR could be increased in the sea- 
sons immediately following PR application by initial application of soluble P fer- 
tilizer (Fig. 6-6a). Crop yield with medium- to high-reactive PR could be 
increased by application of soluble P in later seasons when the residual effect of 
the PR is declining (Fig. 6-6b). The combined use of PR and soluble P fertilizers 
merits agronomic and economic assessment. 

In conceptual terms, a one-time application of soluble P dramatically 
increases capital P, and then the capital P gradually decreases as P is increasing- 
ly sorbed and converted to inert P (Fig. 6-7). A one-time application of a suffi- 

High-reactive PR, 
periodic soluble P - 


ciently reactive PR can dramatically increase capital P, which includes PR that 
releases P to liquid P pools for up to 10 yr. The maximum level of capital P will 
be lower following application of PR than of soluble P, but PR might have longer 
lasting effects on capital P. Applications of PR, particularly sedimentary PRs of 
low reactivity or igneous PRs, would contribute primarily to inert P that does not 
become available to plants within 10 yr and hence is not capital P. The relation- 
ships illustrated in Fig. 6-7 will depend upon reactivity of the PR, soil properties, 
and the cropping system. 

The residual effect of soluble P fertilizer on building soil P capital may 
eventually be greater with seasonal applications than a single large, one-time 
application (Fig. 6-7; Cox etal., 1981; McCollum, 1991; Linquist et al., 1996). 
The decline in extractable P; with time after application of soluble P fertilizer is 
generally a first-order chemical reaction (McCollum, 1991). The decline is par- 
ticularly rapid when extractable soil Pi is high (Cox et al, 1981; Linquist et al., 
1996), as is the case after large P applications. Cox etal. (1981) found for soils 
with contrasting P-sorption capacity that seasonal P applications gave either an 
increase or a more gradual decrease in extractable Pi than a one-time P applica- 
tion. Hence, the seasonal application of soluble P at rates exceeding P uptake by 
plants can gradually build up liquid P and capital P to higher levels than a one- 
time application of P (Fig. 6-7). The gradual build up of capital P from seasonal 
applications rather than an immediate build up with a large P application, how- 
ever, can result in considerably less crop yield for several seasons on moderate 
and high P-fixing soils (Fig. 6-5b). 

Because of the decline in liquid P and capital P after P application, a large 
application of P to rapidly rebuild soil P capital must be followed within a few 
years by periodic maintenance applications of P in order to maintain capital P. 
This strategy of a one-time corrective P application to reestablish soil fertility lev- 
els for high productivity followed by periodic maintenance applications of P, as 

Soluble P or high- 
reactive PR, corrective 


described by Pieri (1987) and Sanchez et al. (1997, this publication) for Africa 
and successfully used in the Cerrado of Brazil (Goedert, 1987; Lopes & 
Guilherme, 1994), will build and maintain soil P capital (Fig. 6-7). It can imme- 
diately eliminate P deficiency and then maintain soil P above deficiency levels, 
thereby resulting in higher cumulative crop yields than with either a sole, one- 
time application of P or seasonal P applications (Fig. 6-5b). The first mainte- 
nance application of P can be delayed until several seasons after the corrective P 
application without loss of crop yield. The length of this delay will increase with 
increasing rate of the corrective P application and increasing P-sorption capacity 
of the soil. 

One-time applications of PR should be compared in agronomic studies with 
seasonal applications of soluble P and corrective plus maintenance applications 
of soluble P rather than with only one-time applications of soluble P. Seasonal or 
periodic applications of medium- and high-reactive PR also merit investigation. 
Application of PR for several years can result in accumulation of undissolved PR 
in soil. The subsequent release of P from this undissolved PR could result in 
appreciable residual value (Rajan et al., 1996). Undissolved PR represents capi- 
tal P if it supplies liquid P (a service flow) within 10 yr. 

Application of Soluble Phosphorus Sources 

Agronomic studies in Africa indicate occasional but not consistent superi- 
ority of spot or band placement compared with broadcast and incorporation of 
soluble P fertilizers (Warren, 1992). Experiences from the Brazilian Cerrado led 
to the recommendation of initial broadcast and incorporation of soluble P fol- 
lowed by annual band applications of soluble P on soils with high P-sorption 
capacity and very low available P (Sanchez & Salinas, 1981). Broadcast applica- 
tion was superior to band application in the season after P application (Yost et al., 
1979). Band application, however, leads to gradual yield increases and greater 
residual effect than broadcast applications (Yost et al., 1981). 

Band or localized placement of soluble P can be effective on P-deficient 
soils with moderate or low P-sorption capacity (Holford, 1989). Fox and Kang 
(1978) found for maize on a sandy loam Alfisol that band application was supe- 
rior to uniform incorporation of TSP at only suboptimal P rates. 

The application of small and moderate quantities of soluble P fertilizer (10 
to 30 kg P ha" ') to maize by either mixing in the planting hole or by broadcast 
and incorporation can be economically attractive on moderate P-fixing, P-defi- 
cient soil (Jama et al., 1997; Table 6-1). In the results from western Kenya (Table 
6-1), added net benefits were negative for application of N without P (Treatment 
1) because the application of 100 kg urea-N ha" 1 without P did not increase yield 
in the first season and slightly decreased yield in the second season. Applications 
of TSP with urea up to 100 kg P ha" 1 increased yield and provided positive net 

Economic Factors 

In conceptual terms, PR will be more attractive than a high-analysis, solu- 
ble P fertilizer when RAE is greater than the relative cost of PR, defined as (farm 


gate + application costs per kilogram P from PR)/(farm gate + application costs 
per kilogram P from soluble P fertilizer) x 100. Even though unreactive PRs can 
have a low relative cost compared with soluble P fertilizer, they are agronomi- 
cally and economically unattractive for direct application because of their very 
low RAE (Fig. 6-8). Sukulu Hills PR, for example, is a large deposit in eastern 
Uganda near P-deficient soils in western Kenya, but it has a very low RAE as 
reflected by its failure to increase maize yields in a pot study (Butegwa et al., 
1996a). Sukulu Hills PR would require processing, such as beneficiation and 
compaction with soluble P, before being used. 

Processing of unreactive PR, such as Sukulu Hills, would dramatically 
increase its RAE (Butegwa et al., 1996a), but the cost of the processed PR rela- 
tive to high-analysis, soluble P fertilizer also would increase. The processed PR 
must have a higher RAE than relative cost to be more economically attractive 
than high-analysis P fertilizer (Fig. 6-8). Phosphate rock processed by partial 
acidulation or compaction with soluble P fertilizer has a lower total P content, and 
hence higher transportation cost, than high-analysis P fertilizer. Therefore, 
processed PR, when compared with high-analysis P fertilizers, will be relatively 
most attractive in agricultural areas near to the processing facility (Hammond et 
al, 1986b; McClellan & Notholt, 1986; Henao & Baanante, 1997). 

Economic analyses have indicated that many PRs are not economically 
attractive for direct application (Seyoum & Mclntire, 1987). Minjingu PR is the 
only known PR deposit in eastern and southern Africa with sufficient quantity 
and reactivity (Table 6-2) for potential in large-scale direct application. The 
deposit is located relatively near western Kenya, where many soils are deficient 
in plant-available P (Kenya Agricultural Research Institute, 1994; Jama et al., 
1997). Minjingu PR is currently not marketed in western Kenya, but an estimat- 
ed retail price of Minjingu PR in western Kenya in December 1996 ranged with- 
in USS 13 to 1.8 kg" 1 P (F. Place, ICRAF, 1997, personal communication). The 

Relative PR cost 


retail price of TSP in western Kenya in December 1996 was about US$ 2.36 kg 
P. The estimated cost of Minjingu PR on a P basis in western Kenya might there- 
fore be about 55 to 76% of the cost of TSP. This value must be interpreted with 
caution. The retail price of Minjingu PR, estimated from acquiring moderate 
amounts of PR for research and a pilot project, could change if the PR were 
mined and delivered to western Kenya on a large scale. 

Labor costs will be slightly higher for application of PR than TSP. PR has 
a lower total P content than TSP, and the fine particle size of ground PR can result 
in considerable dust during handling and application on farms. This dust can lead 
to dislike of the ground PR by farmers and to higher labor requirements for appli- 
cation of ground PR as compared with granular, high-analysis P fertilizer (Ballo, 

Currently available data indicate that the estimated 55 to 76% relative cost 
of Minjingu PR compared with TSP compares favorably with the approximately 
75% RAE of Minjingu PR on acid soils in western Kenya. Bromfield et al. (1981) 
reported a RAE of 75% for Minjingu PR in the five seasons following applica- 
tion to maize in western Kenya. In a recent study with maize on an acid, P-defi- 
cient soil in western Kenya, the RAE in the first season after P application aver- 
aged 74% at 50 kg P ha- 1 and 80% at 250 kg P ha- 1 (ICRAF, 1997). Under such 
conditions of comparable RAE and relative costs, there might be an economic 
incentive for using Minjingu PR in western Kenya. 

Phosphate rocks can contain radionuclides (uranium, radium, and thorium) 
as minor constituents. The concentration of the radionuclides, which depends on 
the origin of the PR, is relatively high in Minjingu PR. This can present an occu- 
pational hazard to workers at the Minjingu mine (Mustonen & Annanmaki, 
1988). Application of P fertilizers at recommended rates has been shown to not 
increase the level of radionuclides in the human food chain. Mortvedt and Beaton 
(1995), for example, reported that annual applications of about 30 kg P ha" ' for 
>50 yr as fertilizer made from Florida PR, which contains radionuclides, had no 
effect on radionuclide concentration in tissues of field crops. The effect of large 
applications of PR, relatively high in radionuclides, on subsequent radionuclide 
concentration in soil components and harvested plant products nonetheless mer- 
its investigation. 

In West Africa, the underutilized medium-reactive Tilemsi PR offers the 
most promise as an indigenous PR source for direct application. Results from a 
5-yr study in four agroecological zones in Mali showed that Tilemsi PR was a 
suitable source of P for sustainable production of important cropping systems. 
Tilemsi PR produced 85 to 90% of the yields obtained with an equal application 
of P from TSP (Henao & Baanante, 1997). Economic feasibility and profitabili- 
ty analyses indicated that the use of Tilemsi PR was financially attractive to farm- 
ers under the prevailing crop prices and climatic conditions. Cost-benefit analy- 
ses indicated that a decrease in PR production from a projected 10 000 t yr" 1 
would substantially increase costs. Poor rainfall and/or a sudden decline in prices 
for crop outputs could make the investment in P unprofitable for farmers, regard- 
less of the P source (Henao & Baanante, 1997). 

In Africa, only a few PRs are presently both agronomically and economi- 
cally attractive to use. Phosphate rocks that are sufficiently reactive for direct 


application often have constraints adversely affecting the economics of use. 
These potential constraints include deposit size, production limitations (e.g., 
source of energy to support mining and grinding operations), lack of infrastruc- 
ture, and in some cases the presence of undesirable constituents (e.g., Cd or 

For the less reactive or unreactive PRs, research on increasing solubility by 
partial acidulation and by compaction with soluble P has provided promising 
agronomic results; however, when all costs (processing, distribution, and market- 
ing) and benefits (increased yields) are considered, it is probable that partial 
acidulation of indigenous PRs will not be the most cost-effective option in large 
regions of Africa. As of 1996, there was no commercial production of PAPR in 

Research has shown that the availability of P from PR can be enhanced by 
composting PR with organic materials (Ikerra et al., 1994; Van den Berghe, 1996) 
and by combination with pyrite, which produces acidity during oxidation (Lowell 
& Weil, 1995). While these small-scale technologies may appear agronomically 
promising, particularly for PR with insufficient reactivity for direct application, 
little is known about tln-u fin in< i il itti ictiven to farmers. 


Soil P deficiency can be corrected either rapidly with investment in a large, 
one-time P application or gradually with seasonal P applications at rates suffi- 
cient to increase availability of soil P. Both approaches will build up capital P. 

One-Time Investment 

A rapid, one-time build up can be achieved with a large application of 
either reactive PR (Fig. 6-5a), soluble P fertilizer (Fig. 6-5a), or low- to medi- 
um-reactive PR combined with soluble P fertilizer (Fig. 6-6a). The P rate 
required to overcome P deficiency will increase with increasing P-sorption capac- 
ity of the soil, just as the residual benefit of the one-time application will be 
longer as P-sorption capacity increases, except for soils with allophane. 

A one-time P application, when integrated with appropriate management to 
overcome other nutrient and crop growth constraints, would ensure a rapid 
increase in crop yields and rapid soil rehabilitation. The increased soil fertility 
could provide an incentive for adoption of higher yielding crop varieties and 
diversification of enterprises within land holdings. This could lead to enhanced 
farm income, which in turn could increase farmer demand for fertilizers to main- 
tain soil fertility. 

Increased crop growth would increase root growth, thereby leading to 
extraction of nutrients and water from a greater soil volume and to increased 
cycling of C in the soil. Increased plant growth could also increase ground cover, 
which could lead to reduced leaching and soil erosion. Application of 500 kg 
TSP-P ha" 1 , for example, to a maize monoculture on an Eutrudox with about 3% 
slope in western Kenya reduced total P loss by erosion and runoff during the 12 


mo following P application (M.R. Rao et al., 1997, unpublished data). The 
reduced P loss was attributed to increased ground cover from weeds and crops 
during the early growth stages of the crop. 

A large, one-time, corrective application of P followed by periodic mainte- 
nance applications of P was an essential component in the successful conversion 
of acid, infertile soils of the Brazilian Cerrado into highly productive soils 
(Abelson & Rowe, 1987). The Cerrado soils were initially high in Al saturation, 
low in exchangeable bases, moderate to high in P-sorption capacity, and very low 
in extractable P (Goedert, 1983). The build up of soil fertility in the Cerrado was 
achieved with large applications of P combined with application of lime or gyp- 
sum, balanced fertilization, improved crop varieties and agronomic practices, and 
an enabling policy environment (Goedert, 1987; Lopes & Guilherme, 1994). Soil 
P capital was built up with relatively large applications of soluble P fertilizers and 
thermophosphates followed by annual maintenance applications of P fertilizers. 

Large, one-time applications of P are proposed as an essential component 
of management for the rehabilitation of degraded acid soils in the humid tropics 
(Dowdle & Von Uexkull, 1988; Von Uexkull & Mutert, 1995). Deforestation of 
acid soils has led to degradation of areas in the humid tropics to anthropic savan- 
na. Anthropic savannas covered by imperata [Imperata cylindrica (L.) Rausch.] 
occupy vast areas in Southeast Asia. As a result of limited nutrient cycling and 
burning, the soils in anthropic savannas are typically low in exchangeable bases, 
high in Al saturation, and low in extractable P. They can be rehabilitated with a 
management package including (i) a large, one-time application of reactive PR (1 
t ha" ') or TSP plus lime (200 kg P ha- 1 and 1 1 Ca ha- 1 ) to correct P and Ca defi- 
ciency and (ii) growth of a leguminous creeper (Mucuna sp.). The legume sup- 
presses regeneration of imperata, provides rapid ground cover, fixes N2 , improves 
water retention, stimulates biological activity, and transforms Pi to P . The large 
application of reactive PR or TSP plus lime ameliorates soil acidity, reduces sorp- 
tion of added P, and increases soil capital P. 

The investment in a one-time P application appears most attractive on high- 
ly P-deficient soils with high P-sorption capacity (Roche et al., 1980). In such 
cases, small applications of P would not markedly increase crop growth in the 
short term. These soils are normally medium to fine textured and frequently high 
in Al saturation. For soils high in Al saturation, the application of P must be 
accompanied by application of Ca, which could require the use of soluble P fer- 
tilizers because dissolution of PR is significantly reduced in the presence of Ca. 

A large, one-time application of P must be viewed as a long-term invest- 
ment because of the long-term residual benefits from the added P. The justifica- 
tion for the investment would be favored by secure land tenure, rapidly escalat- 
ing costs for P inputs, and a low opportunity cost for capital. Moreover, a one- 
time investment that immediately rehabilitates degraded land may in the long 
term be less costly than waiting to reclaim the land after it has been further 
degraded. Large corrective P applications may have particularly large impact 
when targeted to environmentally critical areas (e.g., highly P-deficient soils in 
erosive parts of a watershed). An immediate rather than gradual rehabilitation of 


such areas might provide dramatic environmental benefits (e.g., reduced erosion 
from increased ground cover). 

Gradual Build Up of Soil Phosphorus 

On many P-deficient soils in Africa, relatively moderate applications of 10 
to 20 kg P ha" 1 can dramatically increase crop yields. Such soils normally have 
low to moderate P-sorption capacity and no major constraint from Al saturation. 
Gradual replenishment of these soils could be achieved with seasonal P applica- 
tions at sufficiently high rates to increase the availability of soil P (Fig. 6-5b, 

Seasonal applications of P for gradual correction of P deficiency on soils 
with low to moderate P-sorption capacity will eventually result in greater build 
up of capital P (Fig. 6-7) and greater crop yields (Fig. 6-5b) than a large, one- 
time application of P. Gradual build up of soil P capital, however, will provide 
less immediate and cumulative crop yields than a relatively large corrective P 
application with subsequent maintenance applications ofP on moderate and high 
P-fixing soils (Fig. 6-5b). Seasonal applications will have increased costs (e.g., 
transportation, labor for fertilizer handling and application) compared with either 
one-time applications or corrective plus maintenance applications. 

Gradual replenishment with seasonal P applications would, however, 
enable distribution of a given quantity of P fertilizer to a relatively large land 
area. Replenishment with a large, one-time P application, on the other hand, 
would restrict distribution of the P fertilizer to only a portion of the given land 
area. The remaining land area must wait until subsequent seasons for P applica- 
tion. Gradual replenishment with immediate use of the limited supply of P fertil- 
izer over the entire land area at rates on the steep part of the response curve would 
result in greater aggregate crop yield for the area. Gradual replenishment also 
may provide aggregate environmental benefits (e.g., reduced soil erosion) and 
economic benefits (e.g., increased income) for the area. 

Existing knowledge on immediate and residual effects of P fertilizer (Jama 
et al., 1997; Table 6-1) suggests that the gradual build up of soil P with seasonal 
applications of P can economically increase crop yields on soils with large crop 
responses to relatively moderate P rates (10-20 kg P ha" 1 ). Despite knowledge of 
soil fertility depletion and the need for P fertilizers, many smallholder farmers in 
Africa have not adopted seasonal application of sufficient P for the mitigation of 
soil P depletion. Economic, policy, and infrastructural factors have constrained 
the use of all fertilizers, including P fertilizer. Given this failure of conventional, 
seasonal applications of soluble P to be successfully implemented for mitigation 
of soil P depletion in many smallholder farms in Africa, there is increased need 
to consider corrective plus maintenance applications of P to increase crop pro- 
duction and prevent further environmental degradation. Replenishment with a 
large P application could provide immediate rather than only gradual prevention 
of further soil degradation, even on soils with low to moderate P-sorption capac- 
ity. An excessive rate of P for a one-time application of soils with low P sorption 


capacity, however, might lead to detrimental effects of 
and leaching of P. 

Accompanying Technologies 

Regardless of the option selected for replenishment of soil P, P application 
by itself will not overcome soil fertility depletion. The replenishment of soil P 
must be accompanied by technologies such as soil conservation, integrated nutri- 
ent management, water harvesting, and control of crop pests and diseases to 
ensure the increased and then sustained crop production without environmental 
degradation (see Sanchez et al., 1997, this publication). 

Precautions to reduce erosion and runoff may be particularly important on 
sloping land (Gachene et al., 1997) when combining P application with legume 
rotations for building soil N capital (see Giller et al, 1997, this publication). Van 
Bodegom (1995), for example, found increased soil and P loss by erosion when 
a natural uncultivated fallow was replaced with a planted sesbania fallow in order 
to replenish N fertility on an Eutrudox with 3% slope in western Kenya. Increased 
erosion in the sesbania fallow was attributed at least partly to reduced ground 
cover resulting from removal of weeds during establishment and early growth of 
sesbania. This observation highlights the importance of maintaining soil ground 
cover and surface roughness when replenishing erosive soils. 

In order to increase and then sustain crop production, replenishment of soil 
P must be integrated with replenishment of soil N (Giller et al., 1997, this publi- 
cation) and elimination of other nutrient constraints. Elimination of P deficiency, 
for example, can lead to limitations by micronutrients (Bationo et al, 1995; 
Brodrick et al, 1995). Application of high-analysis P fertilizers without S in con- 
tinuous cropping systems can lead to S deficiencies (Friesen, 1991). Organic- 
based systems and agroforestry do not eliminate the need for P fertilizer inputs, 
but the inclusion of organic-based systems in the replenishment of P can have 
synergism on P availability (Palm et al., 1997, this publication). For example, the 
application of rapidly decomposable organic material to supply N and K can 
enhance plant availability of added P (Le Mare et al., 1987) and reduce P sorp- 
tion (Iyamuremye & Dick, 1996; Ohno & Crannell, 1996). 


There is an indisputable need to correct P deficiency in African soils. This 
can be achieved either gradually with seasonal applications of P fertilizer or 
rapidly with a relatively large, one-time, corrective application of (i) soluble P 
fertilizer, (ii) medium- to high-reactive PR, or (iii) low- to medium-reactive PR 
combined with soluble P. One-time, corrective P applications to reestablish soil P 
fertility must be followed by periodic, relatively small maintenance applications 
of P. 

Only a few PRs in Africa are currently suitable, both agronomically and 
economically, for direct application. Partial acidulation of low- to medium-reac- 
tive PR has not been adopted commerically, and compaction of low- to medium- 
reactive PRs with soluble P remains experimental. Many areas of Africa do not 


have PRs suitable for direct application and will need to rely on imported, high- 
analysis P fertilizers for the foreseeable future. Areas of Africa with PRs suffi- 
ciently reactive for direct application may be able to utilize indigenous PR 

Future assessment of PRs should include the comparison of one-time appli- 
cations of PR with seasonal applications of soluble P fertilizers, including local- 
ized placement of soluble P fertilizer in the planting hole at suboptimal rates as 
practiced by some farmers. The combined application of PRs with soluble P fer- 
tilizers as illustrated in Fig. 6-6 also merits investigation. Sufficient data on costs 
and labor, including extra labor requirements for application of finely ground PR, 
should be collected in order to enable economic analyses. Field experiments 
should be of sufficiently long duration (at least 10 yr) to assess the validity of the 
predicted yield and soil P trends illustrated in Fig. 6-5, 6-6, and 6-7. 

Whereas planted tree fallows, legume rotations, transfer of plant biomass, 
and application of manures have potential to build up soil N capital (Giller et al, 
1997, this publication; Palm et al., 1997, this publication), they cannot suffi- 
ciently build soil P to overcome P deficiency on highly P-deficient soils. 
Integration of organic-based systems and agro forestry with P fertilizers, howev- 
er, can increase P in labile soil P pools and may have potential to enhance the 
availability of soil P. Encouraging results have been obtained in western Kenya 
with the integration of tithonia leaf biomass with P fertilizers (ICRAF, 1997; 
Nziguheba et al, 1998; Sanchez et al., 1997, this publication). 

With regard to soil P, a greater understanding is particularly needed on (i) 
the cycling of P with C and (ii) the mechanisms by which organic materials and 
enhanced soil biological activity influence P availability. Information should lead 
to the development of tools for use in decision making on effective options for 
integration of organic materials and legume rotations with P fertilizers for replen- 
ishment and subsequent maintenance of both soil P and N fertility. 

The replenishment of soil P must be accompanied with management to 
overcome N deficiency and other constraints to crop growth. Approaches to 
building soil fertility must recognize that a major constraint to fertilizer use in 
small landholdings in Africa is low on- and off-farm income. Many smallholder 
farmers have lacked the financial resources to purchase sufficient P fertilizer to 
prevent depletion of soil P. The decision to opt for a large, one-time investment 
in P application followed by periodic maintenance applications of P should be 
based on economic, environmental, and infrastructural factors as well as bio- 
physical factors. 


The authors thank the following for helpful suggestions: CA. Baanante, 
S.H. Chien, B.H. Janssen, P.A. Sanchez, M.J. Soule, J.K. Syers, and S.J. Van 


i agricultural frontier. Science i Washington. DC) 


\ I i i I [ l i I > > 1 1 i i li i f plant 1 i ij 

rus in soils of southern Nigeria. Soil Sci. 122:159-164. 
Adepetu, J.A., andR.B. Corey. 1977. Changes in N and P availability and P fractions in Iwo soil from 
Nigeria under intensive cultivation. Plant Soil 46:309-316. 

\ ! tin i i I 1994. Ph phorus i [iiirt in m o i imii owpea qn n il ci >pping on a Paleudult. 

An I i i I.) i 1996. Phosphat yd in enei rop ems witi inph i i n the availability of 
different pin i phat fra tion ii th I Plant Soil 184:11-21. 

Anderson, G.D. 1970. Fertility studies on a sandy loam in semi-arid Tanzania: II. Effects of phos- 
l hi [i i im i I ii n 11 p. Agri I 

Ankomah, A.B., F. Zapata, S.K.A. Danso, and H. Axmann. 1995. Cowpea varietal differences in 
uptake of phosphorus from Gafsa phosphate rock in a low-P Ultisol. Fert. Res. 41:219-225. 

Aulakh, M.S., and N.S. Pasricha. 1991. Transformations of residual fertilizer P in a semi-arid tropi- 
cal soil undei eight n mi I heat rotation. Fert. 1 14 I 

Baanani ' A 1998. Eo nic evaluation of the f phosphate fertilize pital investment 

/ E. Johnston ai i i i nun ment for sustainabl [ tion in 

i ' I I nt \ llingfonl. Engl mil (in press). 

Baidu-Forson, J., and A. Bationo. 1992. An economic evaluation of a long-term experiment on phos- 
phorus and manure amendments to sandy Sahelian soils: I'sins a stochastic dominance model. 
Fert. Res. 33:193-202. 

Ballo, D.B. 1995. Use of TPR in direct application in the southern Mali region: A factor in sustain- 
able agriculh 1 In H. Gerner and A.I I I fphospl rock foi 
sustainable atiriculture in West Africa. Misc. Fert. Stud. I 1. Int. Fert. Dec. Or.. Muscle Shoals. 

Barrow, N.J. 1983. On the reversibility of phosphate sorption by soils. J. Soil Sci. 34:751-758. 

Bationo, A., E. Ayuk, and A.U. Mokwunye. 1995. Long-term evaluation of alternative phosphorus 
i 1 il n il | i n hi n I i 1 * in ii i i l 

tropics, p. 42-53. In H. Gerner and A. I". Mokwunye led.) L'se of phosphate rock for sustain- 
able agriculture in West Africa. Misc. Fert. Stud. 11. Int. Fert. Dev. Or., Muscle Shoals, AL. 

Bationo, A., S.H. Chien, J. Henao. OB. Christiansen, and A.I '. Mokwunye. 1990 Agronomic eval- 
l lulated II 1 I j I 1 I 

Sci. Soc. Am. J. 54:1772-1777. 

1 ii i nd A.I I 1 in 1991 II tin il fertility nstraints to in I ci i 

duction in West Africa: The experience in the Sahel. Fert. Res. 29:95- 115. 

Bationo, A., S.K. Mughogho, and A. Mokwunye. 1986. Agronomic evaluation of phosphate fertiliz- 
ers in tropical Africa, p. 283-318. In A.U. Mokwunye and P.L.G. Vlek I n 1 

i ill n sub-Sal i Ml jhoff Publ.. Dordrecht 

the Netherlands. 

Beck, M.A.. and P.,-\. Sanchez. 1994. Soil phosphorus fraction dynamics durint! 18 years of cultiva- 
tion on a Typic Paleudult. Soil Sci. Soc. Am. J. 58:1424-1431. 

1 1 I i id tO in ii 199 il i Ii | h hi nun n ni i I mil i iftei 13 years of fertil- 

ized cultivation in the Amazon basin. Plant Soil 1 84:23-3 1 . 

Bekel I nil Hofner. 19 E1 i f dift nt phosphate f rtili i n yield of barley and rape 
seed on reddish brown soils of the Ethiopian highlands. Fert. Res. 34:243-250. 

Bolan, N.S., M.J. Hedley, R. Harrison, and A.C. Braithwaite. 1990. Influence of manufacturing van- 
ii ! in ill ii u In f partial! u 1 i I i I ii i Ii 

Fert. Res. 26:119-138. 

Bowman, R.A., and C.V. Cole. 1978. Transformations of organic phosphorus substrates in soils as 
evaluated by NaHC0 3 extraction. Soil Sci. 125:49-54. 

Braun, A.R., E.M.A. Smaling, E.I. Muchugu, K.D. Shepherd, and J.D. Corbett (ed.). 1997. 
Maintenance and improvement of soil productivity in the highlands of Ethiopia. Kenya. 
Madagascar and I ganda. An inventory of spatial and non-spatial survey and research data on 
natural resources and land productivity. AHI Tech. Rep. Sec 6. African highlands initiative. 
Int. Ctr. for Res. in Agrofor., Nairobi, Kenya. 

Brodrick, S.J., F. Amijee, JA. Kipe-Nolt, and K.E. Giller. 1995. Seed analysis as a means of identi- 
fying micronut nt deficien f ; // L. in the tropi fra] \ n 
(Trinidad) 72:277-284. 

Bromfield, A.R.. I.R. 1 I I L r lam. 198 1 . I t I phosj I 

mental S on yield and P uptake of maize in western Kenya. Exp. Agric. 17:383-387. 


Brouwer, J., and J.M. Powell. 1995. Soil aspects of nutrient cycling in a manure application experi- 
ment in Niger, p. 21 1-226. /;/ J.M. Powell et al. (ed.) Livestock ami sustainable nutrient 

cycling in mixed farming systems of sub-Sahara n Africa. Vol. II. Tech. Pap. Int. Livestock Or. 

for Africa, Addis Ababa, Ethiopia. 
Butegwa, C.N., G.L. Mullins, and S.H. Chien. 1996a. Agronomic evaluation of fertilizer products 

derived from Sukulu Hills. Fert. Res. 44:113-122. 
Butegwa, C.N., G.L. Mullins, and S.H. Chien. 1996b. Induced phosphorus fixation and the effective- 
ness of phosphate fertilizers derived from Sukulu Hills phosphate rock. Fert. Res. 44:231-240. 
< I i m ' i i > u I i hi id EVA LiiKjui ] In] i i u pin in u iii the cumulative soy- 

bean response to pb phori n an Lltisol Id Ci I es. 34:23-36. 
Chauhan, B.S., J.W.B. Stewart, and E.A. Paul. 1981. Effect of labile inorganic phosphate status and 

organic carbon additions on microbial uptake ol phosphorus in soils. Can. J. Soil Sci. 

n, S. H., and L L Hammond 1989 \gron mil i ft ti\ n s if partial!; icidula I pin phatt 

rock as influenced by soil phosphorus-fixing capacity. Plant Soil I 20: 1 59- 1 64 
a, S.H., L.A. Leon, and H.R. Tejeda. 1980. Dissolution of North Carolina phosphate rock in acid 

Colombian soils as related to soil properties. Soil Sci. Soc. Am. J. 44:1267-1271. 
n, S.H., and R.G Menon. 1995a. Agronomic evaluation of modified phosphate rock products. 

IFDC's experience. Fert. Res. 41 : 197-209. 
i II i G. Menon. 19951 i i ng tl ronom i i [I i i 

for direct application. Fert. Res. 41:227-234. 
n, S.H., R.G. Menon, and K.S. Billingham. 1996. Phosphorus a\ lilability from phosphate rock 

as enhanced by water-soluble phosphorus. Soil Sci. Soc. Am. J. 60:1173—1177. 
Cox, FR. 1994. Cunent pin phoru ilability indict Chai (eristics and shoitcomings, p. 

101-1 I 3. In J.L. Havlin and J.S. Jacobsen (ed.i Soil resting: Prospects for improving nutrient 

recommendations. SSSA Spec. Publ. 40. SSSA, Madison, WI. 
Cox, F.R., E.J. Kamprath, and R.E. McCollum. 1981. A descriptive model of soil test nutrient levels 

following fertilization. Soil Sci. Soc. Am. J. 45:529-532. 
i i ',1 hi II bl ii r. 199 v literature i i nil luation of the Hetll fractiona 

i | i i ii ii I i in I In rus in natural in 

Geoderma 64:197-214. 
Dalai, R.C. 1977. Soil organic phosphorus. Adv. Agron. 29:83-117. 

ill il ulii i I i i i in i i I roblem ii liffereix e zone of the devel- 

oping countri ■. ol ub- lahai m Africa, p. 37-52. In H. Van Reuler and W.H. Prins (ed.) The 
1 i 1 nl u in a la nd "I in I i nl ; i h i hi hi I haran Africa ni ng » in 

Kunstmest Producenten. Leidschendam. the Netherlands. 
Delhaize, E. 1995. Genetic control and manipulation of root exudates, p. 145-152. In C. Johansen et 

al. i i Geneti m lip n i ip I nt to nhanct int > , it nut i nt m n . m in in 

cropping systems: 1. Phosphorus. Int. Crops Res. Inst, for the Semi-Arid Tropics. Patancheru. 

Andhra Pradesh, India. 
De Swart, P.H., and A. Van Diest. 1987. The rock phosphate solubilizing capacity of Pueraria javcm- 
i I | i i i I i u i I ii 1 il ii n i i I I nl Soil 

Di, H.J., L.M. Condron, and E. Frossard. 1997. Isotope techniques to study phosphorus cycling in 

agricultural and forest soils: A review. Biol. Fert. Soils 24:1-12. 
1 i m ii i ' 1 i n ii i n t ] u i r i Ii in n p. 448-463. In 

Seminar on direct application of phosphate rock. IFDC Spec. Publ. S-l. Int. Fert. Dev. Or., 

Muscle Shoals, AL. 
Doumbia, M.D., M. Ouattara, L.R. Hossner, A.B. Onken. 1992. Phosphorus status and sorption in 
ml il i H 1 ion i > i i i i ' I i i t i i it i t i I i i u 

Dt i i.n upport ; tern Worl hop nil gt tation TX, 11-12 Mar. 1992. TropSoils Bull. 

92-01 l]i i ! i 1 H 1 Ii i id Human Resour., Univ. of 

Hawaii. Honolulu. 
Dowdle, S.F., and H.R. von Uexkull. 1988 Uanagement practii fo favorabit nutrient cycling in 

tropical upland systems, p. 91-97. /// Proc. of an Int. Conf. on Management ami Fertilization 

of I 'pland Soils in the Tropics ami Subtropics. Nanjing. Peoples Republic tit China 7- 1 I Sept. 

Espinosa. 1. 1992. Phosphorus diagnosis ami recommendations in volcanic ash soils, p. 109-1 15. In 

Proc. of the T op ils Phosphorus D in i'n n tern \ kshop. < lb jr Station. TX. 

11-12 Mar. 1992 [rop I nl ,1 i II 92-01. Dep. of Agron. and Soil Sci., College of Trap. Agric. 

and Human Resour., Univ. of Hawaii, Honolulu. 


FAO. 1996. FAO fertilizer yearbook. Vol. 45—1995. Food and Agric. Organ, of the United Nations, 

FAO/UNESCO. 1977. Soil map of the world, 1:5 000 000. Vol. VI. Africa. Food and Agric. Organ. 

of the United Nations, Rome. 
Fardeau, J.C., G. Guiraud. and C. Marol. 1996. The role ofisotopic techniques on the evaluation of 

the agronomic effectiveness of P fertilizers. Fert. Res. 45:101-109. 

Flacli E.N.. W. Quak. and A. Van Dies'. 1987. A comparison o1 the rock phosphate-mobilizing capac- 
ti f varioi rop sj i I i i ["rinidad 

Fox. R.L.. and E.J. Kamprath. I970. Phosphate sorption isotherms for evaluating the phosphate 
requirements of soil. Soil Sci. Soc. Am. Proc. 34:902-907. 

Fox, R.L., and B.T. Kang. 1978. Influence of phosphorus fertilizer placement and fertilization rate on 
maize nutrition. Soil Sci. 125:34-40. 

Fox. T.R.. and X.B. Comerford. 1992. Influence of oxalate loading on phosphorus and aluminum sol- 
ubility in Spodosols. Soil Sci. Soc. Am. J. 56:290-294. 

Friend, M.T., and H.F. Birch. 1960. Phosphate responses in relation to soil tests and organic phos- 
phorus. J. Agric. Sci. (Cambridge) 54:341-347. 

Friesen, D.K. 1991. Fate and efficiencx o1 sulfur fertilizer applied to food crops in West Africa. Fert. 
Res. 29:35-44. 

Frossanl. E.. VI. Brossard. M.J. Hedley. and A. Vletherell. I995. Reactions controlling the cycling of 
P in soils, p. 107-137. In H. Tiessen led.) Phosphorus in the global en\ ironment:Transfers, 
cycles and management. Sco] i 54 I ihn o ilej & Sons, New York. 

Frossard, E., J.W.B. Stewart, and R.J. St. Arnaud. 1989. Phosphorus distribution as related to its form 
and mobility in grassland and forest soils in Saskatchewan, ("an. J. Soil Sci. 69:401-416. 

Gachene, C.K.K., N.J. Jarvis, H. Linner, and J.P. Mbuvi. 1997. Soil erosion effects on soil properties 
in a highland area in central Kenya. Soil Sci. Soc. Am. J. 61:559-564. 

Gahoonia, T.S., and N.E. Nielsen. 1992. The effects of root-induced pH changes on the depletion of 
i i i i i i i | i i in il ih | I I i i il 1-1 I 'I 

Gerner, FL, and G. Harris. 1993. The use and supply of fertilizers in sub-Saharan Africa, p. 107-125. 
/.<■> SJ. Van Reuler and W. Prins (ed.i The rule of plant nutrients for sustainable food crop pro- 
duction in suh-^ hi i hi frica. Vei nigh n Kim tin t Pn u i t i L id In n I m m 

Gijsman, A. J., A. Oberson, H. Tiessen, and D.K. Friesen. 1996. Limited applicability of the CEN- 
TURY model to highh ■■ i , thered tropical soils. Agron. J. 88:894-903. 

Giller, K.E., G. Cadisch, C. Ehaliotis, E. Adams, W.D. Sakala, and P.L. Mafongoya. 1997. Building 
soil mi n [ i I , i in i p I I ' i i i li il i r i i n i i il fertil- 

ity in Africa. SSSA Spec. Publ. 51. SSSA, Madison, Wl (this publication). 

Goedert, W.J. 1983. Management of the Cerrado soils of Brazil: A review. J. Soil Sci. 34:405-428. 

Goedert, W.G. 1987. Management of acid tropical soils in the savannas of South America, p. 
109-127. In Mai in nt ol nil I i i u in i I n Itui II I \ i li i in I 

Workshop. Int. Board toi Soil J mil in i. in ' ok Thailand. 

Haggar, J.P., G.P. Warren, J.W. Beer, andD. Kass. 199i . Phosphorus availability under alley cropping 
i I il I u i in I I I i l ii i i mi 1 II m! I 

Hammond, L.L.. S.H. C'liien. and G.W. Easterwood. 1986a. Agronomic effectiveness of Bayovar 
phosphate rock in soil with induced phosphorus retention. Soil Sci. Soc. Am. J. 50:1601 -1606. 

Hammond, L.L., S.H. Chien, and A.U. Mokwunye. 1986b. Agronomic value if imacidulated and par- 
tially acidulated phosphate rocks indigenous to the tropics. Adv. Agron. 40:89-140. 

Hammond, L.L., S.H. Chien, A.H. Roy, and A.U. Mokwunye. 1989. Solubility and agronomic effec- 
tiveness of partially acidulated phosphate rocks as influenced b\ their iron and aluminum 
oxide content. Fert. Res. 19:93-98. 

Hammond, L.L., and LA. Leon. 1983 ronom f i t natural and alteted phosphate 

rocks from Latin America, p. 503-518. hi Proc. Int. Congr. on Phosphorus Compounds. World 
Phosphate Inst.. Brussels, Belgium. 

Hands. VI. R.. A.F. Harrison, and T. Bayliss-Smith. 1995. Phosphorus dynamics in slash-and-burn and 
I ley cropp I i I I i T I.) Phosj I 

II 1 t: transl I i i -r I 1 I 


He, Z.L., J. Wu, A.G. ODonnell, and J.K. Syers. 1997. Seasonal responses in microbial biomass car- 
bon, phosphorus and sulphur in soils under pasture. Biol. Fertil. Soils 24:421-428. 

Hedley, M.J., G.J.D. Kirk, and M.B. Santos. 1994. Phosphorus efficiency and the forms of soil phos- 
phorus utilized by upland rice cultivars. Plant Soil 158:53-62. 


Hedley, M.J., J.W.B. Stewart, and B.S. Chauhan. 1982. Changes in inorganic and organic soil phos- 
phorus fractions induced by cultivation practices and laboratory incubations. Soil Sci. Soc. 
Am. J. 46:970-976. 

Heliums, D.T., C.A. Baanante, S.H. Chien. 199 Iterna [nil is 1 itilizeis toi the tropics: 

in agronomi i hi iiii ilu i \ I ' 1 , i In i i ill rropsoil I ■ oru 

1 i n uj it i in V ksliop. Coll i n TX, 11-12 Mar. 1992. TropSoils Bull. 

92-01. Dep. of Agron. and Soil Sci.. College ofTrop. (\gric. and Human Resour., Univ. of 
Hawaii, Honolulu. 

Henao, J., and C.A. Baanante. 1997. Improvement of soil fertility and land productivity through the 
efficient management of indigenous and imported sources of phosphorus in Mali. Int. Fert. 
Dev. Ctr., Muscle Shoals. AL (in press). 

Hinsinger, P., and R.J. Gilkes. 1996. Mobilization of phosphate from phosphate rock and alumina- 
sorbed phosphate b\ the mots of rveurass and clover as related to rliizosphere pH. Eur. J. Soil 
Sci. 47:533-544. 

Hinsinger, P., and R.J. Gilkes. 1997. Dissolution i i | ho ph itt roi k in the rliizosphere of five plant 
pecie i i I'll ii i ' il ii i I 

Hoffland. E. 1992. Quantitative evaluation of the role of organic acid exudation in the mobilization 

of rock phosphate by rape. Plant Soil 140:279-289. 
folford. I.C.R. 19 fl ferenl phate applicatioi i i i lation to phosphate 

sorptivity of soils. Aust. J. Soil Res. 27:123—133. 

Holford, I.C.R. 1997. Soil phosphorus: Its measurement, and its uptake by plants. Aust. J. Soil Res. 
lu . 1991. Eff i u n _ nil cid nions on orption and phyto ii lability ii ils with dil 

ferent mineralogies. Soil Sci. 152:463-471. 

Huffman, S.A., C.V. Cole, and N.A. Scott. 1996. Soil texture and residue addition effects on soil phos- 
phorus transformations. Soil Sci. Soc. Am. J. 60:1095-1101. 

ICRAF. 1996. Annual report 1995. Int. Ctr. for Res. in Agrofor., Nairobi, Kenya. 

ICRAF. 1997. Annual report 1996. Int. Ctr. for Res. in Agrofor., Nairobi, Kenya. 

IFDC. 199(1. Minjingu phosphate rock marketing study — A marketing plan. Int. Fert. Dev. Ctr., 
Muscle Shoals, AL. 

IFDC/CIRAD/ICRAF/NORAGRIC. 1996. PR feasibility case studies: Synthesis report. An assess- 
ment of phosphate rock as a capital investment: Evidence from Burkina Faso. Madagascar. 
and Zimbabwe. World Bank, Washington, DC. 

Ikerra, T.W.D., P.N.S. Mnkeni, and B.R. Singh. 1994. Effects of added compost and farmyard manure 
on P release from Minjintni phosphate rock and its uptake by maize. Norw. J. Agric. Sci. 

1 nd A I ii i[ I \ I99i IS I e of soil ph phate b; quential i i Miction 

of soil properties and added phosphorus. Commun. Soil Sci. Plant Anal. 27:2147-2157. 

Iyamuremye, F., and R.P. Dick. 1996. Organic amendments and phosphorus sorption by soils. Adv. 
Agron. 56:139-185. 
:, A.-M.N. 1997. Ecological economics ot investing in natural resource capital in Africa, p. 
237-251. In R.J. Buresh et al. (ed.) Replenishing sod fertility in Africa. SSSA Spec. Publ. 51. 
SSSA, Madison, Wl (this publication), 
a, B., R.A. Swinkels, and R.J. Buresh. 1997. Agronomic and economic •■•■■, aluation of organic and 

inorganic sources of phosphorus in western Kenya. Agron. .1. 89:597-604. 
isen, B.H. 1993. I ill rated nutrient mai meiit In nst >i nil nd mineral fertil i j 

89-105. ;»H. Van Reuler and W.H. Prins (ed.) The role of plant nutri in ind u i u ibl i 

production in sub liaran Africa. ^ eniging van Kunstmest Producenten eit liendam. th 
isen, B.H., D.J. Lathwell, and J. Wolf. 1987. Modeling long-term crop response to fertilizer phos- 
phorus: II. Comparison with field results. Agron. J. 79:452-458. 

Johnsen, F.H., R. Fernandes, L. Mukurumbira. C. Sukiune. and J. Rusike. 1996. Phosphate rock ini- 
tiative country' case study for Zimbabwe. Ctr. for Int. Environ, and Dev. Studies, Agric. Univ. 
of Norway, As, Norway. 

I i' 2. Principlt for usina rtili i I impro red I rallitii iil in luaiida. E\j \°r'\ 
\.S.R., and R.L. Fox. 1977. Phosphate sorption characteristics of some bench-mark soils in West 
Africa. Soil Sci. 124:370-376. 

'• impi ah E I I '0 Re i In I ffeel i I irg. pj lii i n n pli ;ph >ru < n high phosphotus fixing 
soils. Agron. J. 59: 25-27. 


Kanabo, I.A.K., and RJ. Gilkes. 1987. A comparison 

ments of the dissolution of reacth e phosphatt 
retention. Aust. J. Soils Res. 25:251 A 460. 

Kenya Agncultutal i irch Institute. ] i rtili u i n m ndations Vol. 1-22. Fert. Use 

Recommendation Project. Kenya, Agric. Res. Inst.. Nairobi, Kenya. 

Khasawneh, F.E., and E.C. Doll. 1978. The use of phosphate rock for direct application to soils. Adv. 
Agron. 30:159-205. 

Kirk, G.J.D., M.J. Hedley, and D.R. Bouldin. 1993. Phosphorus efficiency in upland rice cultivars. p. 
279-295. In Papers and reports on the management of acid soils. IBSRAM/ASIALAND 
Network Doc. 6. Int. Board for Soil Res. and Manage., Bangkok, Thailand. 

Kpomblekou, K, S.H. Chien, J. Henao, and W.A. Hill. 1991 Cier i I i i phate 1 

tilizers produced from Togo phosphate rock. C'ominun. Soil Sci. Plant Anal. 22:63-73. 

kiio. S. 1990. Phosphate sorption implications on phosphate soil tests ami uptake by corn. Soil Sci. 
Soc.Am. J. 54:131-135. 

( mi i ml ,1 Hun on loo tral .• ofph pin i .. ui ition ml onseivation by plant 
species and comnuu i p. 139-1 'I n (ed.) horns in th bal environ 

inent: transfers, cycles and management. Scope 54. John Wiley & Sons, New York. 

I 1 P.H. I I [ [I Ul| I 1 11 II I j 1 II il u i I I . In 

D.L. Hawksworth (ed.) Proc. of CAB' s Fu I i ntifi Conf. Advancing Agricultural 
Production in Africa. Arusha. Tanzania. 1 .1- i 8 Feb. 1984. Commonwealth Agric. Bureaux. 
Farnham Royal, England. 

Le Mare. P.H 

able phosphate in a dark-red latosol in Brazil. J. Soil Sci. 38:199-209. 

Linquist. B.A.. P.W. Singleton, and K.G. Cassman. 1997a. Inorganic and organic phosphorus dynam- 
ics during a build-up and decline of available phosphorus in an Ultisol. Soil Sci. 162:254-264. 

Linquist, B.A., P.W. Singleton, K.G. Cassman, and K. Keane. 1996. Residual phosphorus and long- 
term management strategies tor an Ultisol. Plant Soil 184:47-55. 

Linquist, B.A., P.W. Singleton, R.S. Yost, and K.G. Cassman. 1997b. Aggregate size effects on the 
sorption and release of phosphorus in an ( 'Itisol. Soil Sci. Soc. Am. J. 61:160-166. 

Lins, I.D.G., and F.R. Cox 1989. Effect of extractant and selected soil properties on predicting the 
optimum phosphorus fertilizer rate lor urowino soybeans under field conditions. Commun 
Soil Sci. Plant Anal. 20:319-333. 

Lins, I.D.G., F.R. Cox, and J.J. Nicholaides, III. 1985. Optimizing phosphorus fertilization rates for 
soybeans grown on Oxisols and associated Entisols. Soil Sci. Soc. Am. J. 49:1457-1460. 

Loganathan, P., N.O. Isirimah, and D.A. Nwachuku. 1987. Phosphorus sorption by Ultisols and 
Inceptisols of the Niger Delta in southern Nigeria. Soil Sci. 144:330-338. 

Lopes, A.S., and L.R.G. Guilherme. 1994. Solos sob Cerrado Manejo de fertilidade para producao 
i i ii 1 1 i 1 i i 1 i I DA ' t f i I 

Lopez-Hernandez, D., G. Siegert, and J.V. Rodriguez. 1986. Competitive adsorption of phosphate 
with malate and oxalate by tropical soils. Soil Sci. Soc. Am. J. 50:1460-1462. 
o 'i I nd R.R il 1 ' I* i i no i in i yh ph u i til bilit from African phos- 

phate rocks: A laboratory study. Soil Sci. Soc. Am. J. 59:1645-1654. 

Lupin il ml r ' L I'' i i ni[ hi i — n In u ipproach I gi nular fertilizers Tech. 

Bull. T-25. Int. Fert. Dev. Or., Muscle Shoals, AL. 

Mackay, A.D., J.K. Syers, R.W. Tillman, and P.E.H. Gregg. 1986. A simple model to describe the dis- 
solution of phosphate rock in soils. Soil Sci. Soc. Am. J. 50:291-296. 

Marschner, H. 1995. Rhizosphere pH effects on phosphorus nutrition, p. 107-115. In C. Johansen et 
al. led.) Genetic manipulation of crop plants to enhance integrated nutrient management in 
cropping systems: I. Phosphorus. Int. Crops Res. Inst, for the Semi-Arid Tropics. Patancheru. 
Andhra Pradesh, India. 

McClellan, G.H., and A.J.G. Notholt. 1986. Phosphate deposits of tropical sub-Saharan Africa, p. 
173-224. / I I i i i I VI i t i i hosphon 

fertilizers in sub-Saharan Africa. Martinus Nijhoff Publishers. Dordrecht, the Netherlands. 

McCollum, R.E. 1991. Buildup and decline in soil phosphorus: 30-year trends on a Typic 
Umprabuult. Agron. J. 83:77-85. 

Menom R.G.. and S.H. Chien. 1990. Phosphorus availability to maize from partially a, idulated phos- 
phate rocks and phosphate rocks compacted with triple superphosphate. Plant Soil 

Menon, R.G., and S.H. Chien. 1995. Soil i sting foi a ailab c phosphorus in soils where phosphate 
rock-based fertilizers are used. Fert. Res. 41:179-187. 


Menon, R.G., and S.H. Chien. 1996. Compaction of phosphate rocks with soluble phosphates. Tech. 

Bull. T-44. Int. Fert. Dev. Ctr., Muscle Shoals, AL. 
Menon, R.G., S.H. Chien, and WJ. Chardon. 1997. Iron oxide-impregnated filter paper (Pj test): II. A 

ie\ ii [ill in 11 "1 lii > eioecosyst. 47:7-18. 

Menon. R.G.. S.H. Chien. and A.N. Gadalla. 19 I.F'ho 1 il ro ! coin] I with uperpli | ! i t 

vs partially acidulated rocks for beans and rice. Soil Sci. Soc. Am. J. 55:1480-1484. 

Menon, R.G., S.H. Chien, and D.T. Heliums. 1995. Iron and aluminum oxide influence on agronom- 
ic effectiveness of modified phosphate rocks. Soil Sci. Soc. Am. J. 59:1762-1767. 

Mnkeni, P.N.S., J.M.R. Semoka, and D. Mkamba. 1995. Bioavailable phosphorus status in some 
benchmark soils o1 dorogoro districi Tanzania Norv I Agrii lei Suppl 21 1(17-115 

Mokwunve. A. I"., and S.H. Chien. 1980. Reactions of partially acidulated phosphate rock with soils 
from the tropics. Soil Sci. Soc. Am. J. 44:477 A 82. 

Mokwunye, A., S.H. Chien, E. Rhodes. 1986. Phosphate reactions with tropical African soils, p. 
253-281. In A.U. Mokwunye and P.L.G. Vlek (ed.) Management of nitrogen and phosphorus 
fertilizers in sub-Saharan Africa. Martinus Xijhoff Publishers. Dordrecht, the Netherlands. 

Mokwunye. A. I'.. A. (It- layer, and E.M.A. Smaling (ed.). 1996. Restoring and maintaining the pro- 
ductivity of V t Afi in iil y to sustainable development I Fen did. 1 Int. I i 
Develop. Ctr., Africa, Lome, Togo. 

Mokwunye, AU and LL H i II I I sciei I il e in the tiopics, p. 

121-134. In R. Lai and P.A. Sanchez (ed.) Myths and science of soils of the tropics. SSSA 
Spec. Publ. 29. SSSA, Madison, WI. 
I I ' i I i i I I i ii in i [ ii j i 

i n i i l n ii i I in ' I ! i lii n il il 161:4 

Mortvedt. .!..!.. and J.D. Beaton. 1995. Heavy metal and radionuclide contaminants in phosphate fer- 
1 i i 1" i ; 1 1 1 n (ed.) Ph i horn u tl lo il nvi mi in transfei I 

and management. Scop:- 54. John Wiley & Sons, New York. 

Mustonen, R., and M. Annanmaki. 1988. Studies on the radiation exposure of workers in connection 
with processing of the Minjingu phosphate in Tanzania. Suppl. to the Rep. 666/622/87. 
Finni <iii 1 Ii ii ii u i ucleai fei Helsinki. Finl ml 
1 II il Ii i I i II hosphorus fertilization i i ihen relationships to 

maize yield in western Kenya. M.Sc. thesis. Moi I Hi\ .. Eldoret, Kenya. 

Nagarajah. S.. and B.M. Nizar. 1982. Wild sunflower as a izreen manure for rice in the mid country 
wet zone. Trop. Agric. 138:69-79. 

Nziguheba, G., C.A. Palm, RJ. Buresh, and P.C. Smithson. 1998. Soil phosphorus fractions and sorp- 
tion as affected b\ organic and inorganic sources. Plant Soil (in press). 

Ob i i i mcl B i in d 199i < i n ind inimal manui tl io d di >l d i iganic matter effects 

on phosphorus sorption. J. Environ. Qual. 25:1 137-1 143. 

Otani. T.. X. Ae. and II. Tanaka. 1996 Phosphorus (Pi uptake mechanisms of crops grown in soils 
with low P status: II. Significance oforaanic acids in root exudates of pitieonpea. Soil Sci. 
Plant Nutr. 42:553-560. 

> Phosphate sorption characteristics of selected major 

Palm. C.A. 1995. ( nitril nil n i 1 agi n. i> i \ n ' to nuti i. nl i qui t in nl i t u i pped plant ■;. 

Agrofor. Syst. 30:105-124. 
Palm, C.A., R.J.K. Myers, and S.M. Nandwa. 1997. Combined use of organic and inorganic nutrient 
i in ii fertility niaintenan md replenishment I I In R.J. Buresh et al. (ed.) 

Replenishing soil fertility in Africa. SSSA Spec. Publ. 51. SSSA, Madison, WI (this public a- 

Pieri, C 1987. Management of ac id tropical soils in Africa, p. 41-61 VnProc. of Management of Acid 
oil il i i i in il 1 i ill ii 1 B Inaugural I i i i in nagu rn 

.aid Brasilia Brazil. 24 Apr.-:'. May 1985. Int. Board for Soil Res. and Manage.. Bangkok, 

Pushparajah. E., F. Chan, and S.S. Magat. 1990. Phosphorus requirements and management of oil 
palm, coconut nd rtibl 1-408. 1 lion p rem nrs 1 istainabl il me 

in Asia and Oceania. Int. Rice Res. Inst, Manila, Philippines. 

Rajan, S.S.S., and B.C. Marwaha. 1993. Use of partially acidulated phosphate rocks as phosphate fer- 
tilizers. Fert. Res. 35:47-59. 

Rajan, S.S.S., J.H. Watkinson, and A.G. Sinclair. 1996. Phosphate rocks for direct application to soils. 
Adv. Agron. 57:77-159. 


systems: 1. Phosphorus. Int. Crops Res. Inst, for the Semi- Arid Tropics. Patancheru, Andhra 

Pradesh, India. 
I" i i i P nil I ill ii i I' 1 ' ption i ni J rptioi ;i ntit\ in nsity parameters related to 

plant-available soil phosphorus. Soil Sci. Soc. Am. J. 58:405-410. 
Robinson, J.S., and J.K. Syers. 1990. A critical evaluation of the factors influencing the dissolution 

of Gafsa phosphate rock. J. Soil Sci. 41.597-605. 
Roche, P., L Grieve, D. Babre, H. Calba, and P. Fallavier. 1980. Phosphorus in tropical soils: 
i I u i n I I i, i ii n u i hi in i 1 i i I ' rid II li ' Inst.. 


! ' I u 1 r I I I I 1 I II I I II 1 1 | J 1 i 

Sale, P.W.G., and A.U. Mokwunye. 1993. Use of phosphate rocks in the tropics. Fert. Res. 35:33-45. 
Sanchez, P.A., W. Couto, and S.W. Boul. 1982. The fertility capability classification system: 

Interpretation, applicability, and modification. Geoclenna 27:283-309. 
i \ I I I ilencia. and C. I 199 il fertility repleni hment in Africa: A 

concept note. p. 200-207. In S.A. Bteiu I in iini[ i n i h in 

ments in Attica i 1 fri Mexico City. 

Sanchez, P.A., and C.A. Palm. 1996. Nutrient cycling and agroforestry in Africa. Unasylva 185 

Sanchez, P.A., and J.G. Salinas. 1981. Low-input technologies for managing Oxisols and Ultisols in 

Tropical America. Adv. Agron. 34:279-406. 
Sanchez, P.A., K.D. Shepherd, M.J. Soule, F.M. Place, R.J. Buresh, A.-M.N. Izac, A.U. Mokwunye, 

F.R. Kwesiga, C.G. Ndiritu, and PL. Woomer. 1997. Soil fertil ty i pit ni ment in Africa: An 
i ii I I I 1 I I I II 

tility in Africa. SSSA Spec. Publ. 51. SSSA, Madison, WI (this publication). 
i I I 1 I ' i i iii I i 1 1 I 

fixation capacity, p. 471-514. In F.E. Khasawneh ft al. (ed.l Tin- role of phosphorus in agri- 
culture. ASA CSSA. and 111 

Schmidt. J. P.. S.W. Bind, and E.J. Kamprath. 1996. Soil phosphorus d\ namics during seventeen years 
of continuous culm tion: Fracti i i in I il Soc Am. J. 60:1168-1172. 

hi Ii I l I 1 1 n 11 d partiall i 'i i I I | I i i i Ii i icduction cost, and 

use. Tech. Bull. IFDC-T-31. Int. Fert. Dev. Ctr„ Muscle Shoals, AL. 
Schwertmann. U., and A. J. Hei ml m. 199 out peel f fertility a iated with the mineralogy 
of highly weathered tropical soils, p. 47-59. /// R. Lai and PA. Sanchez (ed.l Mvths and sci- 
ence of soils of the tropics. SSSA Spec. Publ. 29. SSSA and ASA, Madison, WI. 
t L i I L i I i p tesponse to phos- 

phate rocks in eastern Africa. ILCA Bull. 29. Int. Livestock Ctr. for Africa, Addis Ababa, 

111,1 I ' ' Ph > i, an ' li i_ in unt oil/ I and 1 itili ed in ultui I soils Soil Sci. Soc. 

Am. J. 49:905-911. 
Sharpley, A.N., J.T. Sims, and G.M. Pierzynski. 1994. Innovative soil phosphorus availability indices: 
in 'in ni ii in ins p 115-142. In J.L. Havlin and J. S. Jacobsen (ed.l Soil test- 
i i | i i 1 1 ii i i i mi i ii i pec. Pul 

Madison, WI. 
Sharpley, A.N., H. Tiessen, and C.V. Cole. 1987. Soil phosphorus forms extracted by soil tests as a 
function of pedogenesis Soil Sci. Soc. Am. J. 51:362-365. 
Id i I I I Industrial mineral nphasis on phosphate rod 1 In D.J. McLaren 

i I I urn i i I nd ill i i ii m i lin i J v ii i i 1 

Smaling, E.M.A., S.M. Nandwa. and [ill. lanssen. Soil fertility in Africa is at stake, p. 47-61. In 
R.J. Buresh et al. (ed.) Replenishing soil fertility in Africa. SSSA Spec. Publ. 51. SSSA, 
Madison, WI (this publication). 
Smeck, N.E. 1985. Phosphorus dynamics in soils and landscapes. Geoderma 36:185-199. 

Smith, J.L. 1994. Cycling of nitrog irough microbial activity, p. 91-120. fa J.L. Hatfield and B. A. 

St wart (ed.) il i 1 i i il iuiht\ Lewis Publ., Boca Raton, FL. 

Smyth, T.J., and PA. Sanchez. 1982. Pho ph I •.' I- li o in < u uid . I ibilit in Cerrado soils as 

affected by phosphorus sorption capacity. Soil Sci. Soc. Am. J. 46:339-345. 
Stewart, J.W.B., and A.N. Sharpley. 1987. Controls on dynamics of soil and fertilizer phosphorus and 
sulfur, p 101 1 I ' 1 Follet et al. (ed il i ,nl f ml i nic matter a i nn il com- 
ponents of production systems. SSSA Spec. Publ. 19. SSSA and ASA, Madison, WI. 


Stewart, J.W.B., and H. Tiessen. 1987. Dynamics of soil organic phosphorus. Biogeochemistry 

Syers J.K., and A.D. Mackay. 1986. Reactions of Sechura phosphate rock ami single superphosphate 

in soil. Soil Sci. Soc. Am. J. 50:480-485. 
Tiessen, H., and J.O. Moir. 1993. Characterization of available P bj sequential extraction, p. 75-86 

In MR. Carter (eel. I Soil sampling and methods of analysis. Lewis Pub!.. Boca Raton, FL. 
Tiessen, H., I.H. Salcedo, and E.V.S.B. Sampaio. 1992. Nutrient and soil organic matter dynamics 

under shifting cultivation in semi-arid northeastern Brazil. Agric. Ecosyst. Environ. 

Tiessen, H., J.W.B. Stewart, and C.V. Cole. 1984. Pathways of phosphorus transformations in soils of 

differing pedogenesis. Soil Sci. Soc. Am. J. 48:853-858. 
Tiessen, H., J.W.B. Stewart, and A. Oberson. 1994. Innovative soil phosphorus availability indices: 
\ i i_i .in phosphorus, p. 143-162. In J.L. Havlin and J.S. Jacobsen (ed.) Soil testing: 

1 I I I I I II 1 I llll il 111-, 111 


Van Bodegoi P I ' Vater. nitrogen n pi c pin u d n i i it in 111 i i li systems and maize 

in western Kem a M thesi ningen Agric. I'ni Va ngen the Netherlands. 

Van den Berghe, C.H. 1996. The effect of Matongo rock phosphate and. urea as < ompared to di-ammo- 

nium phosphate in the com) n i nil I i [ i in the Mu nil i i i 

ofBurundi. Fert. Res. 45:51-59. 
Van der Meersch, M.K.. R. Merckx. and K. Mulongoy. 1993. Evolution of plant biomass and nutrient 
litem in i I i ni i i 1 i In i in two all ropj in osteins, p. 143-154. In K. 

Mulongoy and R. Merckx (ed.) Soil organic matter dynamics and sustain. ihilin of tropical 

agriculture. John Wiley & Sons, Chichester, England. 
Vander Zaag, P., and C. Kagenzi. ! 986. The phosphorus requirements of five consecutive potato crops 

on an Andept in Rwanda. Am. Potato J. 63:121-129. 
Yank uwenbergh I. 1991 . Overview of phosphal i] il in I ist and outheast Africa. Fert. Res. 

Van u i h I 1 < i n I I [I rock produeti ituation I 

/ I [ ■ li in 1 I • i i i ii i n i i i n i i ' ii i i i i n i i ni 

i ill \ i t 1 ill i i I J i 1 i n I li I i I 

Int. Fert. Dev. Ctr., Kandy, Sri Lanka. 
Van Reuler, H. 1996 Nutrient management over extended cropping periods in the shifting cultivation 

system of south-west Cote d'lvoire. Ph.D. thesis. Wageningen Agric. Univ.. Wageningen. the 

Von Uexkiill, H.R., and E. Mutert. 1995. Rehabilitation of anthropic savanna, p. 149-154. In H. 

1 1 ii 1.1 Pin ii hi in til i in ii j u in nm iii yd ml i na m nl j 

54. John Wiley & Sons, New York. 
Walker, T.W., and J.K. Syers. 1976. The fate ofphosphorus during pedogenesis. Geoderma 15:1-19. 
Warren. G. 1992. Fertilizer phosphorus sorption and residual value in tropical African soils. XRI Bull. 

37. Nat. Resour. Inst.. Chatham. England. 
Wielemaker, W.G., and H.W. Boxem. 1982. Soils of the Kisii area, Kenya. Agric. Res. Rep. 922. 

PUDOC-DLO, Wageningen, the Netherlands. 
Wild. A. 1973. Crop responses to fertilizers -limitations imposed b\ soil properties, p. 167-178. In 

FAO/SIDA/ARCN regional seminar on shifting cultivation and soil conservation in Africa, 

Ibadan. Nigeria. 2-21 July 1973. Food and Agric. Organ, ofthe United Nations, Rome. 
Wolf, J., C.T. de Wit, B.H. Janssen, and D.J. Lathwell. 1987. Modeling long-term crop response to 

fertilizer phosphorus: I. The model. Agron. J. 79:445-451. 
I I I I ub Siharan Africa: 

ii ii i ii mil F bilit) Rep World Bank, Washington, DC. 
Yost, R.S., E.J. Kamprath, E. Lobato, and G. Naderman. 1979. Phosphorus response of corn on an 

Oxisol as influenced by rates and placement. Soil Sci. Soc. Am. J. 43:338-343. 
Yost, R.S., E.J. Kamprath, GC. Naderman, and E. Lobato. 1981. Residual effects ofphosphorus 

applications on a high phosphorus adsorbing Oxisol of Central Brazil. Soil Sci. Soc. Am. J. 


Building Soil Nitrogen Capital 
in Africa 

Ken E. Giller, Georg Cadisch, Costas Ehaliotis, 
and Edward Adams 

Wye College 

University of London, England 

Webster D. Sakala 

Chitedze Agricultural Research Station 
Lilongwe, Malawi 

Paramu L. Mafongoya 

SADC/ICRAF Agroforestry Project 

Harare, Zimbabwe 


The dynamic nature of N cycling dictates that soil N capital useful for supplying N 
for plant growth must be equated with short- to medium-term, rolling capital (the month- 
ly or annual salary), rather than long-term reserves (gold in the bank). Thus building of 
soil N capital necessitates a focus on the capacity of soils to store organic matter (the prin- 
cipal reserve of N in the soil) and on management strategies to replenish N reserves. For 
a given climate, the capacity of a soil to store organic matter is directly related to its tex- 
ture through both direct effects of clays adhering to organic matter and indirect effects on 
soil aggregation, which render the organic matter protected from decomposition. Some 
African soils can contain up to 300 to 400 kg N ha" 1 in free mineral form within the top 2 
m of soil, which represents a form of vulnerable capital susceptible to leaching. Large 
amounts of organic inputs are required to build up the soil's capital store of N. The 
approaches most suitable for this are improved legume fallows, legume-grass leys, or min- 
imum-tillage systems. When soils are cultivated continuously it may be impossible to 
build up the soil N capital. Even when the capital store of N is replenished, continued use 
of crop sequences and intercrops with grain legumes and green manures, better integration 
of crops and livestock, and optimal use of mineral fertilizers are essential to ensure 
improved fields are maintained. A. major challenge is to work together -with smallholder 
farmers to find attractive methods for building the N capital of their soils. 

Nye and Greenland in West Africa demonstrated that the total stock of C and N 
in the soil organic matter (SOM), referred to as the soil capital store of C and N, 


declines when soils are cultivated but is gradually restored under natural fallows 
(Greenland & Nye, 1959; Nye & Greenland, 1960). Of all nutrients, N is required 
in the greatest quantity for plant growth, and the capacity of soils to supply N to 
plants is inextricably linked to the amount and nature of the SOM. In this review, 
we distinguish three main forms of N capital. Mineral N (NH 4 -N and NO3 -N) is 
termed vulnerable capital, as it is susceptible to losses and can be equated to cash 
in the pocket. The short- to medium-term capital is the N in the SOM, which is 
mineralizable in the relatively short term (months) and medium term (1 to 5 yr) — 
the monthly or annual earnings. Long-term N capital is essentially the more 
recalcitrant part of SOM that contributes relatively little to N supply within 5 to 
10 yr — gold in the bank. Across much of Africa, soils have come under intensive 
cultivation only recently as human populations have increased, often leading to 
greater net losses of SOM due to reduced organic inputs, the faster rate of SOM 
turnover under cultivation, and increased losses from soil erosion. As a result 
SOM and the supply of N have declined as natural fallows shortened or vanished, 
and the short- to medium-term capital store of N has effectively been used up. 

The purpose of this chapter is to explore the possible strategies that can be 
employed to restore the N capital of soils under cultivation in sub-Saharan Africa 
(hereafter referred to as Africa). There is nothing peculiar about African soils, or 
tropical soils for that matter. The diversity of soils in the tropics is just as wide as 
that found at greater latitudes (Sanchez, 1976; Eswaran et al., 1992), although 
extensive areas of some of the oldest exposed land surfaces occur in Africa, with 
the resulting highly weathered acid soils. What distinguishes African agriculture, 
however, is the predominance of agriculturally based economies within which, 
for a variety of reasons, the use of mineral fertilizers by smallholder farmers is 
very limited (Heisey & Mwangi, 1996). This is in stark contrast to agriculture in 
western Europe and North America, where high productivity is sustained and 
nutrient removal in crops replenished by large inputs of mineral fertilizers or ani- 
mal manures. 

The three principal sources of N for crop production are biological N2 fix- 
ation, organic resources recycled within the cropping field or concentrated from 
a larger area, and mineral N fertilizers. As N is rapidly cycled through the soil and 
in many parts of Africa high temperatures and available water favor rapid decom- 
position, the restoration and maintenance of a capital store ofN in soil is partic- 
ularly problematic. A fundamental question is whether attention must be given to 
increasing SOM contents to improve the efficiency of N use by crops, or whether 
it is more appropriate to manage available organic N sources and mineral N fer- 
tilizers in relation to crop demand without major emphasis on maintenance of the 
SOM. Here we discuss the potential for restoration of soil N capital in relation to 
potential sources of N, management methods, and the ability of soils to store and 
supply N for crop growth. The most striking conclusion is inevitably that there 
are no quick-fix solutions to maintenance of all forms of N capital, or SOM. Any 
proposed interventions must generate cropping systems that are productive, sus- 
tainable, and economically attractive for smallholder subsistence farmers. The 
conundrum is that all restorative technologies for improvement of soil fertility 
without the use of mineral fertilizers involve either import of organic materials 
from surrounding land or allocation of land to produce organic materials. In the 


most densely populated areas, land scarcity prohibits the devotion of land to 
restoration of soil fertility. In such regions methods for replenishment of the 
short-term capital N store in soils will be hard to find without either some other 
form of income generation or short-term assistance to purchase fertilizers or 
direct assistance to compensate for loss of agricultural production, at least in the 
short term. 

Mineral Nitrogen: Vulnerable Capital 

The principal forms of mineral N in soil are NH 4 and NO 3 (NO2 is present 
only as a transient intermediate in nitrification). Ammonium can be held as an 
exchangeable cation on negative charges in soil or as slowly exchangeable (often 
referred to as nonexchangeable) NH 4 . Slowly exchangeable NH 4 is found as an 
interlayer cation in some 2:1 clays, mainly in vermiculite and illites where it sub- 
stitutes for K ions that have a similar ionic radius (Nommik & Vahtras, 1982). 
Release of slowly exchangeable NH 4 is favored by small concentrations ofNH 4 
and K ions, which would tend to occur in heavily cropped soils. The potential for 
N storage as slowly exchangeable NH 4 is therefore limited to soils with sufficient 
amounts of the right clay type, such as the Vertisols (Ayed & Wild, 1983). Jones 
(1973) found evidence for significant amounts of slowly exchangeable NH 4 -N in 
soils poor in SOM from dry savanna regions in West Africa. 

Under aerobic conditions, NH 4 is rapidly nitrified to NO 3 even in acid soils 
of the tropics, which may contain small numbers of autotrophic nitrifying bacte- 
ria (Wild, 1972a). Nitrate is highly mobile in soils and is easily lost by leaching 
(e.g., Pleysier & Juo, 1981), with estimates of 40 to 50% of the mineralized N 
being lost under high rainfall environments of West Africa (Mueller -Harvey et 
al., 1985; van der Kruijs et al, 1988). Leaching of mineralized NO3-N is not as 
rapid as that of NO3-N applied as fertilizer due to the time taken for the NO3-N 
to diffuse to the large pores and channels through which water drains preferen- 
tially (Wild, 1972b). In acid soils leaching of NO3-N is retarded due to retention 
of NO 3 ions by positive charges (Wild, 1972b; Wong et al., 1987). Wong et al. 
(1990b) demonstrated that the delay in leaching of NO3 could be predicted from 
the anion-exchange capacity (AEC) of the soil and that the delay increased with 
depth in a Nigerian Alfisol as the AEC increased. Assuming that half of the sites 
were occupied by NO3, Wong et al. (1990a! estimated that soils with an AEC of 
1 cmol c kg" 1 could hold 140 kg of exchangeable NO3-N ha"' in the plow layer. 
The African soils studied had AECs of 0.06 to 0.3 cmol c kg" 1 , which indicates up 
to 40 kg of exchangeable NO3 -N ha" ' might be held in the plow layer. As strong- 
ly acid soils rarely contain much 2:1 clay it is unlikely that both slowly exchange- 
able NH 4 -N and NO3-N retention could be found to any extent in the same soil. 

In savanna environments with a pronounced dry season, NH 4 -N tends to 
accumulate during the dry season as nitrification ceases at a water potential just 
below the permanent wilting point, whereas mineralization proceeds under slight- 
ly drier conditions (Robinson, 1957). There is a flush of NO3 -N at the start of the 


Fig. 7-1. Mineral N in the surface horizons of a sanch granitic soil at Marondera. Zimbabwe. 
/ith manure added showing the flush ofN at the beginning of the rain: 
:e Grant, 1967). 

rains (Fig. 7-1), commonly known as the Birch effect, due to rapid mineraliza- 
tion of killed microbial biomass and labile organic matter released on drying 
(Birch & Friend, 1956; Birch, 1958; Greenland, 1958; Seneviratne & Wild, 1985) 
and also to rapid nitrification of NH 4 -N accumulated during the dry season 
(Wild, 1972a). Wild (1972b) found that a bulge of N0 3 -N moved slowly down 
the profile during the rainy season (Fig. 7-2) but that leaching to below 120-cm 
depth in a bare fallow with approximately 1000 mm of rainfall was substantial 
only at the end of the rainy season when there was water available for the miner- 
alized NO3-N to diffuse into the effective drainage channels. Several other early 
studies showed that large amounts ofN03-N can accumulate in the subsoil (see 
Wetselaar, 1962). 

Rainfall from start of the st 
157 264 805 930 1069 1199 

183 616 657 973 

5O/0 50A) SOffi 50/0 5O/0 SOU S<V0 SO/0 50/0 50/0 50 

in the surface I m of soil during 2 yrat Samaru, Nigeria (source Wild, 


increasing recalcitrance, and the resulting SOM is a complex mixture of mole- 
cules of plant and microbial origin. 

Recent research by Handayanto et al. (1995, 1997) has highlighted the 
importance of reactive polyphenols in binding proteins. The resulting polyphe- 
nol-protein complexes appear to be very resistant to microbial degradation, 
resulting in poor availability of N for plant uptake from residues that have a nar- 
row C-to-N ratio. It is unclear to what extent complexation of polyphenols and 
proteins is a fast route to SOM formation that could be used to build the long- 
term soil N capital. Even if this were possible, the complexes formed are recalci- 
trant and do not appear to act as slow-release fertilizers. Initial experiments in 
which 15 N-labeled residues with a large protein-binding capacity were added to 
soil indicate that the complexed N is not released over three successive crop 
cycles /Cadisch etal, 1997, unpublished data). 

Analysis of a large database of soils information from Sumatra, Indonesia 
revealed that there was a strong effect of pH on the soil C content; the minimum 
C content was found in the range of pH 5 to 6, soils with pH above or below this 
range tended to have greater soil C contents (Hardon, 1936; van Noordwijk et al, 
1997). The precise mechanisms for greater protection of SOM at low pH are not 
fully understood, but greater amounts of Al in solution and decreased rates of 
decomposition are likely to be involved (van Noordwijk et al, 1997). 

Physical Stabilization 

Feller et al. (1991) showed that the amounts of organic C and N stored in 
West African soils depended on the amount of clay and fine silt-sized particles 
present, and this is shown for a wide range of African soils in Fig. 7-3. In many 
acid tropical soils, notably the Oxisols and Ultisols, the clay-size fraction consists 
mainly of iron (Fe) and Al oxides and in Andisols mainly of allophane. Clay soils 
in general retain more C derived from plant residues in the long term (Jenkinson, 
1977; Amato & Ladd, 1992). Clays are important in direct stabilization of both 
organic molecules in soil (which are primarily microbial metabolites) and the 
microorganisms themselves. The exact mechanisms by which clays bind to 
organic molecules are not fully understood (for a detailed review see Theng, 
1979), but the protective effects of clays are certainly due to close adherence 
between the clays and the organic molecules or the surface of microorganisms. 
Interactions between clays and proteins (and other organic molecules) are strong- 
ly influenced by pH, which affects both the cation-exchange capacity of the clays 
and the organic molecules, for example by changing the conformation of proteins 
(Theng, 1979). Early research showed proteins were readily degraded when com- 
plexed with kaolinite but were resistant to hydrolysis when adsorbed onto mont- 
morillonite (Ensminger & Gieseking, 1942; Birch & Friend, 1956). This led to 
the suggestion that the protective effect of expanding (2:1) clays was due to the 
organic molecules becoming entrapped between the clay layers, which could be 
demonstrated in artificially created, clay-protein complexes in laboratory exper- 
iments (e.g., Pinck et al, 1964). Surprisingly, such complexes were not readily 
found in soils rich in smectite clays, and later research showed that formation of 
interlayer complexes is likely to occur only under specific, rare conditions; i.e., 
in strongly acid soils rich in both smectitic clays and SOM (Theng et al., 1986). 


Despite this, substantial evidence remains that clays afford protection to 
microbial proteins, and montmorillonite clays have a greater protective capacity 
against decomposition than kaolinites (Serensen, 1972). Thus the greater ability 
of the smectitic clays to protect SOM from decomposition appears to be due both 
to their greater cation-exchange capacity (CEC) and the larger available surface 




• Forest 
a Grassland 
O Arable 


1 40- 


• * 

O a £ 


A O 


area for interactions, compared with kaolinitic clays in which the clay layers are 
held together tightly by H bonding. 

From the above discussion we might expect that organic matter is more 
readily degraded in the kaolinitic (1:1) clays, typical of highly weathered Ultisols 
and Oxisols in Africa; however, Motavalli et al. (1995) examined N mineraliza- 
tion from tropical forest soils with a wide range of clay contents and mineralo- 
gies but, with the exception of the Andisols, found no clear distinction between 
smectitic and kaolinitic clays. Andisols are highly weathered volcanic soils that 
have peculiar properties in terms of stabilization of SOM due to the surface inter- 
actions between allophane and organic molecules (Boudot et al., 1988), and they 
characteristically have large SOM contents. 

The extent to which plant residues added to soil are protected by clays thus 
depends on how much of this protective capacity is already saturated (Hassink, 
1996). A large part of the protective effect of clays results not from interactions 
between clay and organic colloids but from better aggregation of soil particles 
(Ladd et al, 1993). Organic residues accumulate within soil aggregates, within 
pores too fine to allow access to decomposer organisms (Tisdall & Oades, 1982), 
and microorganisms are sheltered from predation in pores broad enough to allow 
microbial access but too small for invasion by predatory protozoa and nematodes 
(Elliot et al., 1980). It is not clear to what extent such structural effects on pro- 
tection of SOM operate in some tropical soils such as Oxisols and Ultisols in 
which aggregation is mainly determined by the large amounts of Al and Fe 

Tillage reduces this aggregate-related protection of SOM in both temperate 
(Bauer&Black, 1981; Robinson etal., 1996) and tropical soils (Lai, 1974, 1976). 
This results from breaking up soil aggregates and exposing of organic C previ- 
ously inaccessible and unavailable to decomposer microorganisms rather than 
from improved oxygen availability (Rovira & Greacen, 1957). CO2-C is released 
(Powlson, 1980) and N is mineralized (Craswell & Waring, 1972) as a result of 
soil disturbance in laboratory experiments, and this is in accordance with the C 
losses and reduced N mineralization potentials observed when soils are tilled in 
the field. Losses of mineralizable N exceeded total N losses in various South 
African rainfed soils (Dupreez & Dutoit, 1995), indicating the exposure of a 
weakly stabilized SOM pool following tillage. A significant proportion of the soil 
microbial biomass may be directly killed by soil disturbance (Powlson, 1980) and 
become available to decomposition; however, the largest fraction of the SOM that 
decomposes rapidly following cultivation is nonmetabolic and is directly related 
to decreased soil aggregation. 

Water-stable aggregates may be distinguished by the binding agents 
responsible for their formation into microaggregates and macroaggregates 
(Tisdall & Oades, 1982). Microaggregates have persistent binding agents con- 
sisting of polyvalent metal cations associated with aromatic humic material and 
strongly sorbed polymers and are resistant to cultivation. Macroaggregates 
(mainly > 250 < m diameter), however, comprise microaggregates, plant debris, 
and other organic materials and are held together by transient (microbial and 
plant-derived polysaccharides) and temporary (roots and hyphae) binding agents. 
They are disturbed by cultivation and release C and nutrients (Tisdall & Oades, 


1982; Elliot, 1986). Recently the plant debris encrusted by clays and occluded in 
macroaggregates has been identified as the major fraction of SOM that is exposed 
to decomposition as a result of soil tillage (Oades & Waters, 1991; Golchin et al., 
1994a,b). Although the SOM associated with coarse organo-mineral particles 
shows the greatest losses under cultivation (Tiessen & Stewart, 1983; Zhang et 
al, 1988), SOM that is stabilized by direct adsorption is probably more resistant 
to tillage. 

The identification of pools of SOM that turn over at different rates has been 
a major focus of research. Various attempts have been made to identify and iso- 
late fractions of this SOM, which turnover at different rates, based on size and 
density (Pernet, 1952; Christensen, 1992). These have shown interesting differ- 
ences between fractions that may be used as indicators of differences in turnover 
rates between soils. An isolatable active fraction that represents the majority of 
the mineralizable N is likely to remain elusive as net mineralization comes from 
SOM present in particles of all sizes (Magid et al, 1996). Simulation models of 
SOM turnover all contain several theoretical pools of SOM, and these can be used 
to predict the decline of SOM under cultivation. 


In any discussion of balancing nutrient budgets, there is a need to define the 
scale of the system in relation to inputs and outputs of nutrients. Many studies 
have calculated nutrient budgets at the scale of an individual field, but recogni- 
tion that virtually all farming systems depend on nutrient transfers made between 
fields within a farm, to the farm from the surrounding communal land, between 
farms, between villages or regions (e.g., Smaling et al., 1993, 1997, this publica- 
tion) has led to an increased emphasis on nutrient budgets at the farming system 
or watershed scale. At a coarser scale, national nutrient budgets have been cal- 
culated for many African countries (Stoorvogel et al., 1993), and international 
trade results in major transfers of nutrients between continents (Cooke, 1986). In 
this review, we largely consider the N economy at the scales of the individual 
farm, including the catchment area for those farms where nutrients are effective- 
ly being harvested from the surrounding lands. 

African farming systems are diverse with many different carbohydrate sta- 
ple crops, such as yams (Dioscorea sp.), cassava (Manihot esculenta Crantz), 
bananas and plantains (Musa sp.) and sweet potato [Ipomoea batatas (L.) Lam.], 
generally in the more humid areas, apart from the widespread cereal crops maize 
(Zea mays L.), sorghum [Sorghum bicolor (L.) Moench], and millet [Pennisetum 
glaucum (L.) R. Brown]. Nutrient transfers are of particular importance in mixed 
crop and livestock systems where animal manure is a major soil amendment, and 
such systems dominate many savanna regions in West and southern Africa (e.g., 
Prudencio, 1993). Soil erosion and redeposition may represent a lateral transfer 
of nutrients between fields, or between farms in different positions in the land- 
scape, or may indeed represent a loss of soil and nutrients to river and lake sedi- 
ments (van Noordwijk et al., 1997). Soil erosion control s 


retaining SOM, and root systems of crops and trees may also capture nutrients in 
subsurface lateral water flow (Breman & Kessler, 1995). Other transfers away 
from crop land are made by certain gallery-building termites, and termitaria are 
often spread in fields by farmers to improve the productivity of their soils (Wood, 


Mineral Inputs 

Mineral N is added to the soil in fertilizers or through atmospheric deposi- 
tion. The typical patterns of restricted mineral fertilizer use in African agriculture 
described by Bekunda et al. (1997, this publication) occur despite potentially 
large crop yield responses to moderate fertilizer additions. Where production is 
maintained by the use of fertilizers, which are generally applied to cash crops, the 
returns of N and organic materials to the soil in crop residues also can be signif- 
icantly increased (see below). Animal manures also may contain significant 
amounts of N in mineral forms (Grant, 1967). In poorly buffered or acidic soils, 
repeated use of most types ofN fertilizers in the absence of other measures to 
maintain soil pH leads to problems of reduced production due to increasing soil 
acidity (Djokoto & Stephens, 1961; Bache & Heathcote, 1969; Jones, 1976; 
Pichotetal., 1981). 

It is ironic that atmospheric deposition of N in dust or rainfall tends to be 
restricted except in close proximity to the major industrial centers, of which there 
are few in Africa, as the most important source of combined N in the atmosphere 
is gaseous pollution from transport and industry. The limited information avail- 
able indicates that between 0.5 and 12 kg N ha" 1 are contributed on an annual 
basis from atmospheric deposition in Africa (Jones & Bromfield, 1970; Pieri, 
1992), and typical inputs are likely to be at the bottom of this range. Aeolian 
deposits are especially significant in West Africa where the Harmattan winds 
result in the redistribution of large amounts of surface soil, although the inputs of 
organic N have been estimated at only 1.2 to 4 kg N ha" 1 (Herrmann, 1996). 

Organic Inputs: Amounts and Quality 

Traditional Fallows and Crop Residue Management 

Under natural forest or savanna vegetation, an equilibrium content of SOM 
is reached that is related to the amounts of organic material added to the soil, the 
rate of turnover, and the capacity of the soil to retain SOM. Traditionally the 
burning of vegetation after clearance results in loss of much of the N in the litter 
and plant biomass, depending on the intensity of the fire. Once land is opened for 
cultivation, the SOM declines to a new equilibrium content related to reduced 
amounts of organic inputs, the faster rate of organic matter turnover caused by 
tillage, and increased losses of SOM due to erosion. 


After soil fertility has declined, a long period of natural fallow is required 
to restore the SOM to its original content (Nye & Greenland, 1960). Restoration 
of the original N content is due to a number of factors including the lack of 
removal of N in harvested produce, the gradual concentration ofN into the sur- 
face SOM due to reduced mineralization rates in the absence of cultivation, 
greater inputs of more recalcitrant OM, N deposition from the atmosphere, and 
uptake of N by deeper rooting species, resulting in more efficient capture and 
recycling of N from deeper horizons. Free-living or root-associated N2 fixation 
generally contributes only small amounts of N in agriculture but may be impor- 
tant during long natural fallow periods when availability of C-rich substrates may 
support higher rates of N2 fixation in the order of 10 to 20 kg N ha" 1 yr 1 (Giller 
& Day, 1985). 

In agricultural fields, N may be recycled back into the soil in crop residues, 
both above and below ground, but these are often insufficient to maintain the 
SOM and the N supply at an adequate content for productive agriculture. In cer- 
tain traditional practices, poor-quality plant litters are collected from surrounding 
land. For example, in the fundikila system in northern Zambia poor-quality 
organic material is collected from a large area and composted in situ in mounds 
of soil both to provide nutrients for crop growth and to reduce effects of toxicity 
and P fixation in the very acid soils. This has been likened to recycling poverty 
as such cultivation is backbreaking work that often results in pitifully small yields 
(Dudal & Deckers, 1993). More commonly, approaches to soil N management 
require inputs of N-rich organic material, such as legume residues and animal 

Grain Legumes as Sole Crops and Intercrops 

The role of leguminous crops in maintaining soil fertility is well recognized 
but also has too frequently been uncritically overestimated. Tropical grain 
legumes can certainly fix substantial amounts ofN (Table 7-1) given favorable 
conditions, but the majority of this N is often harvested in the grain. Legumes 
such as soybean [Glycine max (L.) Merr.] that have been subject to intense breed- 
ing efforts are very efficient at translocating their N into the grain, and even when 
the residues are returned to the soil there is generally a net removal of N from the 
field (Halvin et al., 1990; Peoples & Craswell, 1992; Giller et al., 1994). Some 
promiscuous soybean varieties are leafier, have a greater potential to add N to the 
soil, and are potentially more appropriate for cultivation by smallholder farmers 
than the recommended varieties grown on commercial farms in southern Africa 
(Mpepereki et al., 1996). Soybean residues at harvest are lignified (-10% lignin) 
with C/N ratios around 45:1 and these tend to immobilize N when they are added 
to the soil (Toomsan et al., 1995). By contrast, groundnut (Arachis hypogaea L.) 
residues can contain >160 kg N ha"' , are less lignified (-5% lignin), and are rich 
in N, as the crop is harvested while still green. If returned to the soil, groundnut 
residues can easily lead to doubling of maize yields on sandy soils (McDonagh et 
al., 1993), but even with groundnut there is a net contribution from N2 fixation 
only if the legume stover is returned to die soil or if substantial leaf fall occurs 
before harvest. The exceptions to this rule are the longer duration grain legumes 

to soil ill i i n [ i i i i i i the tropics in parent! > 




N 2 



Grain legume 

Duration yield 








Arachis hypogaea 

Sole crop 

90-140 0.8-2.7 






(106-119) (0.3-3.1) 





(105) (0.8) 



Cajanus cajan 

Sole c rop 



(140-241) (1.1-1.4) 









Glycine max 

Sole crop 





(97-104) (2.4-3) 




(0 A 50) 


Net input 


kgNha" 1 

(114) (1.6-4) 





(70-90) (0.5-1.5) 



Sole crop 

69-115 0.2-1.4 














: Balasubramanian and Nnadi, 

0-87) (-37—46) (14-23) 

8-73) (59-73) 

l;MacColl, 1989; McDonaghetal., 1993; Suwanarit etal., 1986; Toomsanet al., 1995; Wetselaar and Ganry, 
1982; 2: Dalai, 1974; Jones and Wild, 1975; Kumar Rao and Dart, 1987; MacColl, 1989; Cobbina, 1995; Mandimba, 1995; 3: Wetselaar and Ganry, 1982; Suwanarit 
et al., 1986; Ofori and Stern, 1987; MacColl, 1989; Sisworo et al., 1990; Toomsan et al., 1995; 4: Jones and Wild, 1975; Davis and Garcia, 1983; Davis et al., 1984; 
Ssali and Keya, 1984b, Ssali and Keya, 1986; Castellanos et al., 1996; Amijee and Giller, 1998; Gilleret al., 1998; 5: Agboola and Fayemi, 1971; Balasubramanian 
and Nnadi, 1980; Eaglesham et al., 1982; Ssali and Keya, 1984a; Ofori et al., 1987; Van Kessel and Roskoski, 1988; Ntare et al., 1989; Sisworo et al., 1990; Bationoo 
etal., 1991; Franzluebbers etal., 1994; Klaij et al., 1994; Reddy et al., 1994. 


such as pigeonpea [Cajanus cajan (L.) Millsp.] and varieties of cowpea [Vigna 
unguiculata (L.) Walp. sp. unguiculata], which may lose a substantial amount of 
biomass in the form of roots and leaves that fall before harvest (Giller & Cadisch, 
1995). A sole pigeonpea crop drops up to 40 kg N ha" 1 in fallen leaves during its 
growth (Kumar Rao et al., 1983), and its small harvest index means that a rela- 
tively large proportion of the fixed N remains in the field (Table 7-1), which can 
give a substantial benefit to subsequent crops. But virtually all the information 
that we have on contributions from N2 fixation is from research conducted on 
experimental stations where the crops have been adequately fertilized with P and 
other nutrients, and often irrigated (Giller & Wilson, 1991). 

As biomass and yields of sole-cropped grain legumes under smallholder 
conditions in Africa are often small (<500 kg ha" 1 of grain), the amounts ofN2 
fixed are barely significant. For example, in the Usambara Mountains in northern 
Tanzania, where bean (Phaseolus vulgaris L.) is the staple grain legume, most 
farmers' crops lacked nodules because of severe P deficiency, and amounts of N2 
fixed were estimated to be as little as 2 to 8 kg N ha" 1 (Amijee & Giller, 1998; 
Giller et al., 1998). Adding fertilizers to alleviate the P deficiency resulted in sub- 
stantial enhancement of nodulation; however, only when K fertilizers were also 
supplied were grain yields raised to 1000 kg ha" 1 or more at most sites (Smithson 
et al., 1993), which would result in roughly 50 to 60 kg N ha" 1 from N2 fixation. 
Amounts of N2 fixation by grain legumes also can be severely constrained by 
drought, and Ganry (in Wetselaar & Ganry, 1982) found that N2 fixation by 
groundnuts over three years in Senegal was almost linearly correlated to total 

Intercropping of grain legumes generally results in the legume deriving a 
greater proportion of its N from N2 fixation than when grown alone, but legume 
dry -matter production and N accumulation are usually reduced because of com- 
petition from the companion crop (e.g., Nambiar et al, 1983) so that the overall 
amount of N2 fixed is less (Table 7-1). Cowpea intercropping was advantageous 
when intercropped with maize or millet in seasons with adequate rainfall, but the 
cowpea competed strongly with the cereal crop for soil water when rainfall was 
limiting (Shumbaetal., 1990; Franzluebbers etal., 1994). One notable exception 
again is pigeonpea, which has a phenology complementary to that of most cere- 
al crops. As the initial above-ground growth and development of pigeonpea is 
very slow, there is little direct competition between the crops (Dalai, 1974). The 
long duration of traditional pigeonpea varieties and their ability to root deeply 
allow the pigeonpea to grow on after the companion cereal crop has been har- 
vested, utilizing residual water in the soil; however, although sole pigeonpea gave 
clear residual effects on growth of subsequent maize, the residual effects of 
maize-pigeonpea intercrops were not substantial (Kumar Rao et al, 1983, 1987), 
presumably because of reduced inputs of N. Despite claims for substantial trans- 
fer of N for grain legumes to companion cereal crops, the evidence indicates that 
benefits are limited and largely due to sparing effects (Giller et al., 1991). 
Benefits are more likely to accrue to subsequent crops as the main transfer path- 
way is due to root and nodule senescence and fallen leaves (Ledgard & Giller, 


Improved Fallows: Green Manures and Agroforestry Species 

Direct benefits from N2 fixation are obviously greater when herbaceous or 
shrubby legumes are grown specifically to improve soil fertility as green manures 
or planted/allows (Tables 7-2 and 7-3). Amounts of N accumulated by the 
legume are generally determined in the short term by the rate of establishment of 
the legume, and subsequently by the productivity of the legume and the length of 
the growing period for the green manure or planted fallow. 

Early reports of experiments successful in maintaining crop yields using 
green manures resulted from testing of a rather limited selection of species such 
as mucuna [Muciina prut tens var. utilis* (L.) DC], pigeonpea, Crotalaria sp., and 
Canavalia sp. across a wide range of environments (e.g., de Sornay, 1918; Davy, 
1925). Green manures such as sunnhemp (Crotalaria juncea L.) were exploited 
extensively to maintain soil fertility on commercial farms in Zimbabwe until min- 
eral N fertilizers became widely available (Rattray & Ellis, 1952). The dense 
cover formed by creeping legumes such as mucuna and kudzu [Pueraria phase- 
oloides (Roxb.) Benth.] leads to self-shading and senescence of leaves giving a 
dense mat of organic matter. Inputs of N from such species based solely on mea- 
surements of standing crop may be underestimated substantially (van Noordwijk 
& Purnomisidi, 1992). Significant N benefits in yields of subsequent crops have 
been reported even when mucuna was burned to ease land preparation (Vine, 
1953), supporting the suggestion that large amounts of N were contributed to the 
soil from roots and fallen leaves. 

Early examples demonstrated the successful use of planted green manure 
fallows in restoring soil fertility more rapidly than regeneration of the native veg- 
etation (e.g., Jaiyebo & Moore, 1964). It also was quickly recognized that 
although green manures gave greater yields of subsequent crops than rotation 
with grain legumes such as groundnut, "they suffer the handicap of occupying the 
land unproductively for a whole year" (Brown, 1958). Thus additional benefits 
such as improved weed control or other uses are generally necessary for farmers 
to spontaneously adopt use of green manures (see below). There has recently 
been a resurgence of interest in the use of short-term, planted fallows using shrub- 
by legumes such as sesbania [Sesbania sesban (L.) Merr] and tephrosia 
(Tephrosia vogelii Hook.f.), with demonstration of substantial gains in crop 
yields (Kwesiga & Coe, 1994; Sanchez etal, 1997, this publication). 

Intercropping with Green Manures or Trees 

Intercropping and relay cropping of legume green manures have the advan- 
tage that crops are still produced while organic material is produced for soil 
amendment. The obvious disadvantages are that the green manures or trees may 
compete with the crops, and that the amounts of organic material produced are 
generally less than when the land is devoted to soil improvement (Tables 7-2 and 

' The taxonomy of Mucuna is contusing ami there are main species in this pan-tropical genus, but 
the nonstinging varieties used in agricultural experimentation invariably belong to this species (R. 
Polhill & B. Verdcourt, 1996, personal communication: Wulijarni-Soetjipto & Maligalig, 1997). 

Residual effect 
X contributed Recovery of in fertilizer 

below ground stover N equivalents References 

Anichis repens 


Cajanus cajan 


b) 2.1-3.7 



I 'tilapogonium muciiiwklcs 



b) 4.1-5.7 



Canavalia ensiformis 


9 b)4.8 
b) 3.4-5.1 


( eiiiroxemaaciitifoliiim 






Centrosema macrocarpum 

(120-1 yr) 





l 'i'i!lrn.\cinti ptixciioriun 



Centrosema pubescens 

100-1 yr 



67- 1 36 

99- 1 30 




Cliaiiiaccrishi rotiiiuliftkia 



Crotakirkt juncea 

_'U- ( J(i 




Crotalaria ochroleuca 



Desmodhim canum 





Desmodium heteropln II urn 

i y r 

b) 1.5-3.2 


De.smotlilllll nvtilifnliillll 






Table 7-2. Continued. 

Residual effect 


N from N 2 

Amount of 


N contributed 

Recovery of 

in fertilizer 

Green manure legume 




N 2 fixed 



stover N 



d unless stated 

tha" 1 



kg N ha" 1 

- kg N ha ' 

Macroptilium atropurpureum 


b) 2.5-5.0 





Mimosa pudica 


b) 1.3-1.8 




Mucuna pruriens var. utilis 




5, 8, 9, 12, 

Neonotonia wightii 






Pueraria phaseoloides 

100-1 yr 


b) 4.1-5.7 








Psophocarpus palustris 






Sesbania cannabina 







Sesbania rostrala 





Sesbania sesban 







Stylosanthes capitata 








Stylosanthes guianensis 

(120-1 yr) 

a) 2-10 






Stylosanthes hamata 





7, 14, 23, 24 

Stylosanthes macrocephala 








b) 6.8-8 



Vicia saliva 


Vigna luteola 


b) 1.2-2.9 




Vigna radiata 





Vigna unguiculata 






Zomia glabra 








\ 1: Agboola and Fayemi, 1972b; 2:Agboola and Fayemi, 1972a; 3: Bartholomew et al., 1992; 4: Cadisch et al., 1989; 5: Cahn et al., 1993; 6: Cobbina, 1995; 7: de 
Leeuwetal., 1994; 8:Drechseletal., 1996; 9: Fischler, 1997; 1 0: Gethin Jones, 1942; ll.Grof, 1986; 12; Hairiah, 1993; 13: Johnetal., 1992; 14:Kouameetal., 1993; 
15:KwesigaandCoe, 1994; 16:Meeluetal„ 1994; 17:Morrisetal., 1986; 18: Mulongoy andAkobundu, 1985; 19: Pateletal., 1996; 20: Reynolds, 1982; 21: Sanginga 


kg N ha" 

soil fertility by shrubs and tr 


X from 

N 2 

input equivalent 

kgNha" 1 


Acacia refwiens 



Acacia tarlilis 



( 'alliantlra calt>th\i:\ii\ 





5, 13, 14 

Casuariiia cqiliseli folia 




8, 15 

Dichrostachys cinerea 



ErxUirina poeppi^iana 




Faidherbia albida 


6. 12, 17 

Flciiiingia macraplirlla 




10, 19 

Gliricidia sephnn 






13, 14, 18, 19 

Lcilcacna Icilcncepliaia 





43- 1 83 

4,5,7, 10, 15, 
3. 13 

Parascriaulhes fai cat aria 


43- 1 83 

Pericopsis angolensis 




Prosop is glandulosa 




128 A 175 


2, 3, 5, 10, 19 

Sesbania formosa 


1 8-32 


Sesbania grandiflora 




4, 5, 13 

Sesbania rostrata 






Sesbania sesban 



60- 1 34 


1: Comet et al., 1985; 2: Danso and Morgan. 1993b: 3: Danso and Morgan. 1993a; 4:Duguma et 
al„ 1988; 5: Duguma and Tonye, 1994; 6: Dunham, 1989; 7: Haggar et al., 1993; 8: Ladha et al., 
1993; 9; Matthews et al., 1992; 10; McDonagh et al„ 1995; 11: Mwiinga et al., 1994; 12: Okorio 
andMaghembe, 1994; 13: Peoples et al., 1991; 14: Peoples et al., 1996; 15: Sangingaet al., 1986; 
16: Sanginga et al., 1989; 17: Schulze et al., 1991; 18: Tian et al.. 1993: 19: Yamoah et al., 1986; 
20: Yoneyama et al. 1990. 

7-3). Whether intercropping with green manures and trees is advantageous thus 
depends on the balance between the benefits and the costs. 

The net benefits may vary significantly between sites and seasons, depend- 
ing on the availability of water and nutrients, and the unpredictable nature of the 
interactions between the green manure and the crop adds a risky complication. 
Relay planting can reduce the likelihood of competition with the crop where rain- 
fall is limited, with the production of the green manure restricted by its ability to 
use residual water after the main cropping season. Hedgerow intercropping or 
alley cropping has been very useful for developing a better understanding of 
tree-crop interactions (i.e, see van Noordwijk, 1996; Vanlauwe et al., 1996), but 
its applicability in smallholder agriculture still remains to be demonstrated 
because of strong crop-tree competition and the intensive management required 


(Young, 1989; Sanchez, 1995). An exception may be on steeply sloping lands, 
where hedgerows can be planted on contours to help prevent soil erosion. 

The traditional agro forestry practice of farmers who maintain trees such as 
faidherbia [Faidherbia albida (Del.) A. Chev.] in their fields is well documented 
as a means for maintaining fertile islands of soil around the trees (Dancette & 
Poulain, 1969; Vandenbeldt, 1992). The extent to which this practice actually 
develops rather than maintains soil fertility is still unclear. Similar effects are seen 
under the canopies of other N2-fixing trees such as Acacia species and also under 
trees that cannot fix N 2 (Belsky et al, 1993; Breman & Kessler, 1995), and the 
extent to which N2 fixation contributes to this phenomenon needs detailed inves- 
tigation. Root systems of trees in arid lands can scavenge for water by being very 
extensive, extending up to 50 m from the trees in some species (Soumare et al., 
1993) or by rooting deeply. The potential for intensification of tree planting with 
species such as faidherbia to enhance the soil N status will depend on the relative 
importance ofNi fixation orN acquisition from a wide area in the enhancement 
of soil fertility (Giller & Cadisch, 1995). If there is significant N2 fixation that is 
deposited on the soil surface through leaf fall then current natural stands of trees 
could be intensified substantially. 

Animal Manures 

Cattle manure is an integral component of soil fertility management in 
many regions of Africa. The beneficial effects of manure on soil fertility are well 
documented, and crop responses to manure application are often due more to the 
contribution of P and cations such as Ca and Mg than the addition of N (Hartley, 
1937; Grant, 1967) or due to physical effects of SOM addition on water infiltra- 
tion and retention (Mugwira & Murwira, 1997); however, crop responses to 
manure application observed in farmers' fields are highly variable due to differ- 
ences among farmers and between regions in the chemical composition of the 
manures, in the rates of manure application, and in the frequency of application 
on each field. 

The nutrient contents of manures differ due to variation in the animals' diet 
and in particular due to differences in the ways manure is collected, supplement- 
ed, and stored. In regions with a long dry season, the quality of grazing available 
is often much better during the rains or after fire, resulting in a larger N content 
in the manure. Powell (1986) found that dry-season manure had an N content of 
6 g kg" 1 of dry matter compared with 18.9 g kg" 1 during the early rainy season, 
when the quality of diet had improved. The diet also may influence the partition- 
ing of N between feces and urine; feeding high-quality diets (containing little 
lignin and polyphenols) results in more N being excreted in the urine than in the 
feces (Reed et al, 1990; Somda et al, 1995). Feeds rich in tannins increased the 
amount of N excreted in the feces as compared with urine, where N is often 
quickly lost through volatilization. Recent results indicate that the N in manures 
from animals fed with tannin-rich diets is very resistant to mineralization in soil 
(P.L. Mafongoya, 1997, unpublished data). 

Animal management has an important influence on the amount of manure 
collected and its N content (Table 7 — 4). There are two principal systems for 


11 from different animal n 





Burkina Faso 







Grazing in fields 



Confinement of herds 
fields overnight 














Transfer/communal ar 





South Africa 



Cote d'lvoire 

1: Campbell er af. 1997: 2: Grant. 1967: 3: Hartley. 1937: 4: Karigwindi et al 
1972; 6: Murwira, 1995; 7: Powell and Mohamed-Saleem, 1987; 8: Powell an 
Prober! et al., 1995; 10: Quilfen and Milleville, 1983; 11: Rodel et al., 1980. 

manure collection from animals: Those in which animals are penned continuous- 
ly and the manure is collected, stored, and transported to the cropped fields, and 
those where animals are allowed to graze in grazing areas during the rains and 
graze freely during the dry season but are corralled at night. In some situations 
animals are corralled on the cultivated field between harvest time and time of cul- 
tivation of the next crop. This system is better economically in terms of manure 
transport, and the effects on soil fertility are generally greater because of the 
inputs from urine. As cattle that are corralled often graze extensively over large 
areas, the collection of manure represents an enrichment of fertility from a wide 
area onto the field where the manure is used. Storage of manure before applica- 
tion to the field has been shown to influence the N content of the manure. 
Ammonia may be lost rapidly by volatilization from manures (Murwira, 1995). 
When manure was stored in heaps or in pits until application, the buried manure 
had substantially greater contents of N, P, and K (Kwaye, 1980). Addition of crop 
residues or straw to manure reduced N losses, and there is certainly scope to 
improve manure management to enhance its value in supplying nutrients in syn- 
chrony with crop demand. In many regions, however, there are strong demands 
to use or sell crop residues for fodder or as building materials. 

Surprisingly, the beneficial effect of manure on N availability for maize 
grown in granitic sandy soils was due to N released directly after application (Fig. 
7-1). This N was most likely present as free mineral N in the manure as miner- 
alization studies have shown that the poor-quality manures found in Zimbabwe 
lead to a prolonged period of N immobilization (Murwira & Kirchmann, 1993). 
Yield responses to manure can be seen in crops for several years after application 
when the manure is supplied in sufficiently large amounts (Mugwira & Murwira, 


Grain legumes 

Fast-growing tret 


Row 'strip intercropping Crop n 

Row 'strip intercropping Green 

Residue transfer 

Grass/arable leys Stall-feeding/corralling 

Sandford (1989) estimated that 16 to 47 ha of grazing land were required 
to produce sufficient manure for sustained maize production of 1 to 3 t ha" 1 in a 
semiarid environment in West Africa. It is clear that there is insufficient manure 
to sustain even such moderate yields in many parts of West Africa (Fernandez- 
Rivera et al, 1995; Williams et al., 1995). There also is a danger of long-term 
degradation of grazing lands, as there is substantial nutrient removal over pro- 
longed periods. In Burkina Faso farmers' rates of manure application were mea- 
sured at 2.5 to 41 ha" 1 yr 1 (Quilfen & Milleville, 1983). Considering cattle con- 
fined in bomas, Probert et al. (1995) calculated that manure available in 
Machakos, Kenya was sufficient to supply cropland with only 2.5 t ha" ' annual- 
ly although much larger rates in excess of 38 t ha" 1 were applied by farmers to a 
few fields. In a sandy soil in Niger a large part of the N applied in manure was 
translocated to depths below 15 m after application of 13 t ha" 1 of manure, indi- 
cating that smaller, more frequent applications may be a more effective way of 
using manure (Brouwer & Powell, 1995). Thus although manure is an important 
source of nutrients for crop growth in many cropping systems in Africa, it is 
widely acknowledged that insufficient manure is available to support crop pro- 

Provision of Organic Inputs 

In summary, there are three basic ways of producing organic inputs rich in 
N for use in soil fertility improvement: crop sequences or fallows with grain 
legumes or green manures; simultaneous intercropping systems where crops and 
green manures are grown together; or biomass transfer or cut-and-carry systems. 
Animals also play a role in converting N-poor crop residues into somewhat rich- 
er sources of N. All of these methods for producing biomass can be used with 
either herbaceous green manures or with fast-growing trees (Table 7-5). Apart 
from the cut-and-carry systems, N is added to the soil from leaves that fall dur- 
ing growth, from shoot material returned to the soil when the plants are harvest- 
ed, and from roots and root exudates. The amounts of N returned below ground 
are difficult to quantify and therefore very poorly documented, but often they rep- 
resent the only input of organic residues to the soil. The major problem with all 


these different approaches to providing organic resources is the limited quantity 
available unless a substantial investment of land, and labor and other n 
committed to their generation. 

Can We Build Soil Nitrogen Capital? 

Given the susceptibility of mineral forms ofN (the vulnerable capital) to 
losses by leaching and gaseous losses, strategies for building soil N capital must 
be focused on the short- to medium-term N capital in the SOM. As discussed 
above, the capacity of a soil to store organic matter (referred to as the equilibri- 
um level by Nye & Greenland, 1960) is determined largely by the soil's texture 
and soil pH (van Noordwijk et al, 1997). This maximum capacity to store organ- 
ic matter relates to the amount of organic residues, microbial biomass, and 
microbial metabolites that can be stabilized. While this storage capacity may be 
exceeded by applying extremely large amounts of organic inputs, rapid turnover 
of unprotected organic matter (including turnover of the microbial biomass itself, 
Ladd et al, 1995) will lead to loss of this excess organic matter unless it is con- 
tinually replenished. The relationship between clay and silt content and the soil C 
and N contents is illustrated in Fig. 7-3. In clay-rich soils the amount of C and N 
is highly variable, depending on the quantity of organic inputs (often related to 
rainfall) and the land use and in particular on the intensity of cultivation. The soil 
C and N content under forest soils can be much greater than soils under grassland 
or arable cultivation because of surface accumulation of unprotected organic mat- 
ter in the surface horizons. Sandy soils invariably contain a small amount of C 
and N, irrespective of the land use, because of their lack of capacity to protect 
organic matter from microbial degradation. Therefore, the degree to which the 
organic matter content (and hence the soil N capital) can be built up depends on 
how much of the protective capacity of the soil is already saturated (Nye & 
Greenland, 1960, p. 53; Hassink, 1995b). 

The positive effects of SOM on crop growth are many (e.g., increasing 
porosity, infiltration, resistance to erosion, ease of root penetration) and do not 
solely depend on the capacity to supply N or other nutrients (de Ridder & van 
Keulen, 1990). In the granitic sandy savanna soils of Zimbabwe, which have 
roughly 5 g kg" 1 organic C and a N content below 0.4 g kg- 1 , Grant (1967) con- 
cluded that SOM was a minor source of N for maize growth and that supplemen- 
tation with mineral fertilizers or manure was essential to ensure reasonable yields 
even in the short term. 

Application of large amounts of N-poor residues, highly lignified residues, 
or residues rich in polyphenols may allow accumulation of amounts of SOM 
above the clay-determined storage capacity, but this will lead to relative enrich- 
ment of the SOM with chemically recalcitrant, passive pools. As such pools must 
be relatively inert to allow them to accumulate, it is unlikely that they can con- 
tribute much directly to N availability for crops, although there will be other ben- 
efits from the effects of increased SOM on soil structure. 


Literature on tropical soils abounds with statements that suggest that SOM 
contents cannot be replenished under cultivation because of thesis/ rate of oxi- 
dation. There are examples where SOM has been increased, depending on the 
amount of organic residues returned to the soil. For example, Bache and 
Heathcote (1969) demonstrated that the addition of cattle manure at 2 t ha" 1 for 
15 yr led to small increases in soil C (from 2.4 to 4.3 g kg' 1 ) and N (from 0.21 to 
0.34 g kg' 1 ) contents at Samaru, Nigeria, and Pichot et al. (1981) found that annu- 
al applications of 60 t ha' 1 of cattle manure increased soil C from 2.5 to only 6.6 
g kg" 1 after 18 yr at Saria, Burkina Faso. At Kabete, Kenya, addition of 10 t ha" 1 
of cattle manure combined with return of all crop residues failed to prevent a 
decline in the SOM contents in an Alfisol cropped annually to maize and beans 
(Kapkiyai, 1996; Smaling et al., 1997, this publication). 

The potential effects of different cropping patterns and residue manage- 
ment on equilibrium SOM contents are illustrated by modeling exercises based 
on the data of Siband (1974) for millet cropping systems in Casamance, Senegal 
(Fig. 7—4 and 7-5). The simulations indicated that return of crop residues had a 
negligible effect on the amounts of soil C (Fig. 7-5a). Rotation with groundnut 
where all stover was returned to the soil had only a small impact on soil C but 
helped to maintain better millet yields in the alternate years in which it was 
grown. Addition of 1 t ha" ' of cattle manure increased production, but to main- 
tain soil C at contents close to those found under the forest, annual application of 
5 t ha" 1 of cattle manure was required: much more than is available in most 
African cropping systems. The simulations suggest that the addition of 45 kg N 
ha" 1 annually to the millet gives an initial boost in amounts of soil C, but that this 

Fig. I -A. Changes in soil organic n 
Soils (pH 6.1, clay 12 to 14%) 
vation for different lengths of til 

tier with length ol cultivation in a reel soil in Casamance, Senegal. 
ere sampled from profiles in five fields that had been under culti- 
- ranging from undisturbed forest (0 yr) to 90 yr. 


amount of N is insufficient to maintain production and organic inputs, as the ini- 
tial SOM is turned over more rapidly with cultivation. A grazed ley system with 
the pasture legume stylo [Stylosanthes guianensis (Aublet) Sw.; which has shown 
potential in such regions, Tarawali & Peters, 1996)] maintained a higher equilib- 
rium C because of the relatively lignified residues, the longer growing period, and 
a small proportion (10-25%) of the fixed N exported in animal products (Fig. 
7-5b). If the land was cropped to millet for 50 yr before groundnut and complete 
crop residue retention were introduced (Fig. 7 -5c), the increases in soil C were 
much less than where these changes were initiated from the start (Fig. 7 -5 a). This 
indicates a hysteresis between SOM depletion and restoration because of the poor 
growth of crops when the soil becomes badly degraded and SOM contents are 
substantially reduced. Although these interventions are simulated and do not 
include the likely effects of increased soil erosion when the soils become degrad- 
ed, this exercise illustrates the need for combined approaches to increasing N 
inputs using legumes, fertilizers, and animal manures if available. 

Managing Nitrogen-Poor Crop Residues 

Crop residues poor in N, such as the cereal stovers, are the major sources 
of organic materials produced in most smallholder food production systems in 
Africa and therefore are arguably an important resource for maintaining the 
organic matter contents of soils. Such residues are often burned to aid plowing 
and assist in pest control. 

Because of their wide C/N ratio and relatively large amounts of C that are 
readily available for microbial growth, a prolonged immobilization of N in the 
microbial biomass is induced, which deprives crops of available N during the 
early growing season. Thus although recycling cereal stover to cropped lands 
may help to maintain SOM contents, or increase them in degraded soils with the 
associated benefits of improved soil structure, the short-term N supply must be 
managed to allow productive cropping. This can be done by use of mineral fer- 
tilizers or by addition of other organic resources rich in N, but sufficient amounts 
of readily available N must be added to satisfy the immobilization potential of the 
cereal straws and allow production of both grain and stover. After 3 yr of incor- 
porating millet straw, a significant increase in soil N mineralization was observed 
and there was a significant increase in the soil N content (Pichot et al, 1974), 
although the amount of straw applied (10 t ha" 1 ) was equivalent to 3 yr of actual 
straw production in the system, assuming that all of the straw was returned to the 
soil. Key questions are how long a period is required before a net benefit is seen 
after cereal straw incorporation and how much straw should be incorporated. 

An alternative approach is to compost all cereal stover or feed it to animals 
to avoid problems of N deficiency in crops sown soon after residue incorporation. 
Both composting of crop residues and feeding to animals help to improve the 
quality of soil amendments and hence the ease of handling as the nutrients are in 
a more concentrated form; however, composting is labor intensive, involves N 
losses, and is more likely to be feasible for maintaining productivity in home gar- 


Minimum or zero tillage are further ways by which the SOM store can be 
increased under intensive cropping. Lack of tillage generally leads to a greater 
equilibrium SOM content because of better conservation of organic residues 

M - GN + St ha' 1 yr" 1 manure 

M - GN + 1 t ha'' yr'' manure 
M - GN rotation 
M - no straw removal 
M - 50% straw removal 

il at Casamance. Senegal, 

' i ii| ing in ii ig hi in i i ii imp i I ill. ii ii i i i i in 1,1 millel in no i ipping i mi nil 
stover returned or SO',- straw removal, millet-groundnut (GN) rotation with all residues returned and 
without cattle manure or with cattle manure at the rates of I or 5 t ha'"; (b) the millet monocropping 
with 50% straw lenu 1 n I th lill i uiidnut I rotation ith 'I In returned incl I i 

no i iin - rt hown t'oi nnpari m together with imulation run for monocropping of millet 
fertilized with 45 kg N ha ' andalyrmill I yr stylo i in i, > "lazed ley rotation; 

i i ui r millet monocropping ith 50'., i iw emo I. folio lb millet monocropping uh ill 
residues returned and the millet-groundnut (GN) rotation with, or without 1 t ha" 1 cattle manure. 
Modeling exercises were done using the CENTURY model (Version 4.0) where soil organic matter 
(SOM) dynamics arc principal!) based on a three-pool model (active pool = 2.\ microbial biomass C; 
slow pool = appioxii i I ii i I I | llli lli tides a nutrient lim- 

ited feedback mechanism on crop growth ami stover production where nutrient demands (defined as 
crop-specific maximum C/N ratios of young tissue) are not matched by N mineralization or inputs of 
X. Initial SOM content was obtained b\ using local climate ami soil conditions under a simulated for- 
est situation until stead} state conditions were reached (=200 yn. Thereafter a i baring year was intro- 
duced with plowing and soil in ition activiti i ui a I m model was run with different 
cropping options for 90 yr. All crop plantings were preceded by plowing. I'nless otherwise indicated 
all crop residues were left in the field in the case of millet and groundnut or 50% were removed for 
alternative uses. \\ h nanui i uli i Ii i i i i ie lant n h ii i 
the millet-stylo ley system the legume was grazed lightly (assuming no effect on subsequent growth) 
three times (manure was deposited in the field). The large effect of the stylo on SOM build up is due 
to (i) the longer growing period than that of the crops, (ii) lack of straw removal, and (iii) a higher 
lignin content (lignin in stylo was allowed to vary from 10 to 25', i n Iin n plant age whereas 
lignin in die crops varied from 6 to I2' : < ). 


within the field, greater physical protection of residues due to the lack of cultiva- 
tion, and reduced losses of SOM through erosion. This gradually leads to an 
increased SOM and soil N content, which is achieved through reduced rates of N 
release. Several experiments in Alfisols of West Africa have demonstrated greater 
SOM and total N contents in soils under zero tillage compared with cultivation 
after only 2 to 4 yr (Kannegieter, 1968; Lai, 1974, 1976). Yields of maize and 

M- stylo grazing 

M - gn + 1 1 ha" 1 yr' 1 manure 
M + 45 kg N ha' 1 yr" 1 
M - 50% straw removal 

M-GN + ithaV 

M - GN rotation 

M - 50% straw removal 


legumes were similar in untilled and cultivated plots to which recommended rates 
of N fertilizers were added (Lai, 1974, 1976). No comparisons were made in 
these studies on the effects of reduced tillage without mineral fertilizer inputs, 
and weed control was achieved by using herbicides in untilled plots (Lai, 1976), 
representing an additional external cost. Dalai (1989) found increases in soil C 
and N contents in the top 10 cm of a fine-textured Vertisol in tropical Australia 
after 13 yr of zero tillage when all residues were returned but the increased N con- 
tents were marked only when mineral fertilizers were applied; however, on anoth- 
er Vertisol small differences in soil C and N contents were found between zero 
tillage and conventional tillage after 8 yr in only the surface 2.5 cm, even with N 
fertilizers and return of all crop residues (Dalai et al., 1995). Similarly, Nyborg et 
al. (1995) achieved a net addition of N to the soil after 11 yr of barley (Hordeum 
vulgare L.) cropping only when fertilizer N was applied to increase plant biomass 

Thus a substantial improvement in SOM content will be necessary before 
the net benefits due to mineralization from the larger amount of SOM outweigh 
the reductions in the net amounts of N mineralization. Other inputs, such as using 
N fertilizers or legume cover crops and herbicides for weed control, also will be 
required. In fact, large benefits in crop production are likely to be found only if 
full tillage is periodically reintroduced after a long period of reduced tillage, and 
tillage operations can be used as a strategic way of mining accumulated N for 
crop production. 


Increasing Nitrogen Inputs 

Unfortunately there are no miracle cures for the restoration or 
of soil fertility. The use of mineral fertilizers by smallholder farmers is common- 
ly restricted (see other chapters in this publication), and used alone, mineral fer- 
tilizers can lead to (or exacerbate) problems of soil acidification unless corrective 
measures also are taken (e.g., Kwaye et al., 1995; Bekunda et al, 1997, this pub- 
lication). Organic resources also are generally in limited supply. The above dis- 
cussion indicates that substantial amounts of N-rich organic materials are 
required to make a significant impact on crop yields. 

All methods for generation of organic material for soil amendment depend 
on allocation of land, potentially resulting in a loss of crop yield. Only where 
marginal land that cannot be used for agricultural production land is used, or land 
is used at a time when no crops could be grown, is there no penalty in lost yields. 
Even when land is used for grazing, not enough is available to produce the 
amounts of animal manure required to sustain crop production. To date there is 
limited use of forage legumes or ley farming systems in African smallholder agri- 
culture that would both increase the fodder available (and hence both the quanti- 
ty and quality of manure) and contribute to soil fertility directly. 


In some parts of Africa, due either to very dense human populations or the 
incidence of trypanosomiasis, cattle are only a minor component of agriculture 
and manure is not an option for crop fertilization. In such situations fallowing or 
green manuring are the most suitable methods for production of organic inputs. 
There are abundant examples in the scientific literature where green manures 
have been shown to be useful for maintaining soil fertility (e.g., Jaiyebo & 
Moore, 1964; Kwesiga & Coe, 1994), although use of such practices by farmers 
is limited. The reasons for lack of adoption of green manures are complex, 
although the extra labor involved in managing a green manure and the unpre- 
dictable responses are often cited as reasons. Very often the necessary stage of 
farmer experimentation and testing of green manure technologies has never been 
carried out, which might have allowed for farmer innovations and modification 
of technologies. Smaling and Fresco (1993) calculated that 50% of the arable 
cropping land in one district of Kenya would have to be devoted to green manur- 
ing to balance the N outputs. 

Where land is very scarce, the intercropping of green manures and cereals 
may be the only feasible means for generating organic inputs. In this case the 
interplay between the reduction in cereal yield caused by competition from the 
intercropped green manure and the residual benefit in yield from the contribution 
of N returned to the soil in the green manure biomass is critical. For a significant 
effect on cereal yields in a subsequent year the absolute minimum green manure 
biomass required would be in the order of 40 kg N ha" 1 (roughly 21 ha" 1 dry mat- 
ter). As approximately 20% of the N from a high-quality green manure residue is 
recovered by the first crop (Giller & Cadisch, 1995), this is likely to give a yield 
benefit of only 500 kg ha" ' . Given that the farmer has had to wait for a year to 
realize this extra yield, then the yield loss that has to be tolerated in the first sea- 
son is less than that returned in the following season. A longer-term residual ben- 
efit from the green manure may compensate for the initial competition with the 
main cereal crop, but as long-term benefits from addition of high-quality organic 
materials tend to be small it is often hard to quantify such gains. There are exam- 
ples where intercropped or relay-cropped green manures produced 81 ha" 1 of bio- 
mass (Table 7-2) although whether such inputs are achievable under farmers' 
conditions remains to be tested. 

Similarly, a 2-yr planted fallow may give treble the yield of continuous 
maize in the same season, but the cumulative maize yield during 3 yr gives vir- 
tually no advantage (Kwesiga & Coe, 1994), although residual benefits of the fal- 
low may extend for two to three subsequent crops. Unless there is no shortage of 
agricultural land for crop production, the yield benefit after a green manure fal- 
low must be substantially greater than that which could have been grown in three 
consecutive crops. But if land is abundant, there is perhaps little incentive to 
increase the yields from a given area. 

Improving Nitrogen Recycling Efficiency in Cropping Systems 

Apart from the N conservation that can be achieved by recycling crop 
residues to the soil, there is substantial scope for increasing the efficient capture 
of mineral N. Much emphasis has been focused on the concept of enhancing the 


synchronization of N release from organic resources with crop demand for N, and 
this is undoubtedly important in very wet climates with high-potential leaching 
risks; however, in the seasonally dry climates that prevail over much of Africa, 
large amounts of mineral N are present in soils at the onset of the rains. Wild 
(1972b) emphasized the need for early growth of crop roots to capture the N 
released in this flush. The common practice among farmers of planting with the 
first rains could partly be in recognition of this, and simple interventions such as 
avoiding deficiencies of other major nutrients could help in ensuring good root 
growth. Planting of crops on ridges can also help to reduce N leaching as much 
of mineralized or added N is in the raised topsoil, and the majority of leaching 
occurs due to water collecting and entering the soil in the furrows (van 
Noordwijk, 1989; Itimu, 1997). 

Obviously the deeper that roots penetrate later in the season the more 
NO3-N can be recaptured from those depths. Losses of N from leaching were 
reduced to almost nothing under mixed perennial crops compared with maize 
under a high-rainfall climate, mainly because of the constant presence of a deep 
rooting system (Seyfried & Rao, 1991). In annual cropping this could be mimic- 
ked through relay cropping with pigeonpea or other deep-rooted cover crop 
species that may capture free NO3-N from deeper horizons. The N could then be 
returned to the soil in the legume residues at the beginning of the next growing 
season. Soils with an appreciable anion exchange capacity tend to be highly 
leached and dominated by Al, which may form an effective chemical barrier to 
root penetration by many crops or trees; however, the soils in western Kenya have 
appreciable NO3 retention capacity, but Al saturation is below 10%, and fast- 
growing agroforestry trees such as calliandra {Calliandra calothyrsus Meissner) 
and sesbania have substantial root length at depths well below those reached by 
maize (Mekonnen et al., 1997; Jama et al., 1998). Amounts of NO3 -N and water 
were significantly reduced in the subsoil compared with those found under maize 
monoculture, suggesting that the trees were effective in capturing N03-N that 
would otherwise have been leached (Hartemink et al, 1996; Mekonnen et al., 
1997). Such trees could therefore be effective ways of reducing N losses (partic- 
ularly where there is subsurface lateral water flow) and returning N to the crop- 
ping system if forms of management sufficiently attractive to smallholder farm- 
ers can be found. 

It also is possible that substantial amounts of mineral N may occur in deep 
water tables, which are accessible to very deep-rooting (tree) species in arid cli- 
mates (Buresh & Tian, 1997). Roots offaidherbia and some Acacia species have 
been shown to reach water at depths of up to 40 m (Dupuy & Dreyfus, 1992) but 
the amounts of mineral N that occur at such depths are not well documented. 

Combined Use of Organic and Mineral Resources 

Initial results examining the interaction between mineral N and organic 
residues using l5 N-labeled urea indicate that there is no bonus to be gained in 
terms of increased fertilizer N use efficiency by mixing mineral and organic 
inputs at moderate rates of addition (Itimu et al., 1997, unpublished data). Only 
when sufficient organic material has been added to have significant effects on soil 


Structurere, nutrient retention, and root penetration is there likely to be increased 
efficiency in the use of mineral fertilizer (Palm et al., 1997, this publication I. 


Large Additions of Organic Matter or Repeated Inputs? 

Fallowing, whether natural or planted, is really the only way to enhance the 
SCM capital store of N, as other sources of organic materials are in limited sup- 
ply. If large amounts of organic manures are added to soils there is a danger that 
the protective capacity of the soil may be exceeded, leading to wastage due to 
rapid decomposition and losses of the mineralized N by leaching, volatilization, 
or denitrification. Further, if the capital store of N in SOM is built up, there will 
always be the temptation to spend [rum the bunk by tillage or liming to stimulate 
N mineralization and release. 

Simply improving the efficiency with which nutrients are recycled within 
existing cropping systems cannot alone give the increments in SOM or soil N 
supply necessary to raise crop production to respectable yields. Repeated addi- 
tions of high-quality organic residues or mineral fertilizers or both are necessary 
to increase and maintain crop production. Indeed, a prerequisite for efficient 
nutrient cycling is a sufficiently deep and dense rooting system to ensure that 
available N is captured. A certain degree of soil fertility is necessary to ensure 
this, and where crop yields are already poor, inputs of N and other nutrients into 
the cropping system are necessary to enhance early crop growth and allow cap- 
ture of free NO3-N at depth. 

The urgency of solutions to meet the growing demand for increased pro- 
ductivity dictate that extra N must be brought into the cropping systems, and this 
can be provided through the use of fertilizers or through addition of fixed No. No 
single solution can, or should, be recommended, as interventions need to be tai- 
lored to, and developed jointly with farmers with due regard to their wide diver- 
sity of farming systems, cultures, and needs. The diversity of soils and soil fertil- 
ity status between fields within single farms or villages also must be recognized 
and exploited to allow gradual implementation of soil fertility restoration strate- 

Critical Targets for Future Research 

Long-Term Soil Fertility Experiments 

The number of long-term experiments in which there is adequate detailed 
information on the effects of continuous organic residue applications on soil N 
build up is still limited for different climates and soils in Africa. Carefully 
planned and managed trials on land representative of soils in farmers' fields are 
necessary to establish the long-term effects of new interventions on crop produc- 
tion and soil fertility (Greenland, 1994). Most experimental stations were estab- 
lished and remain on the more fertile soils, and much of the information cited in 
this review is derived from them. Experiments on more representative soils are 


required to give a much more secure basis for modelling exercises (e.g., Fig. 7-5) 
to explore the likely effect of different management on the capital store of N in 

Farmer-Focused Research 

As we have argued earlier (Giller & Cadisch, 1995), there is perhaps suffi- 
cient understanding of both the processes and the resources that are available for 
the enhancement of soil fertility to make an impact. Yet mere is sparse evidence 
for use of this knowledge in smallholder farming in Africa, at least in part 
because improvement of soil fertility requires a substantial investment of 
resources (whether financial for the purchase of mineral N fertilizer or in terms 
of land and labor for the growth or collection of organic manures). The recent 
resurgence of interest in the use of cover crops such as mucuna in Benin does not 
represent a new intervention: this had been tried with success by experimenters 
in West Africa much earlier, who expressed dismay at the lack of implementation 
by smallholders (e.g., Dennison, 1959). The adoption of mucuna green manuring 
by smallholders indicates that the idea was highlighted by researchers at a time 
when it fitted farmers' needs well. Indeed the ability of mucuna to establish and 
grow quickly and smother imperata grass [Imperata cylindrica (L.) Rausch.] has 
been the main reason for farmers' interest (Versteeg & Koudokpon, 1992). By 
contrast, the direct fertilizer effect of mucuna was the primary benefit highlight- 
ed by farmers in Honduras although they also recognized the advantages of the 
reduced labor requirements for weeding and improved water conservation 
(Buckles, 1995). Rehabilitation of imperata-infested land has also been the main 
reason for recommending mucuna in Southeast Asia, where die predominance of 
acid, P-fixing soils means that large doses of phosphate rock were required 
together with mucuna to allow productive cropping for several years (von 
Uexkull & Mutert, 1994). 

Better mechanisms are required for sharing of knowledge among all of 
those involved — farmers, researchers, extension agents, and NGO workers — in 
trying to improve productivity of smallholder agriculture. Many of the agricul- 
tural researchers and extension officers in Africa (including authors of this chap- 
ter) come from small village communities, and the gap of understanding is not as 
wide as we often imagine. Flow of knowledge cannot be achieved through a nar- 
row prescriptive approach but requires development and testing of a battery of 
possible interventions for soil fertility improvement suited to the specific agroe- 
cological environment together with farmers. Enhanced networking among soil 
fertility researchers in Africa represents a significant recent development [e.g., 
the African Network for Biological Management of Soil Fertility (AfNet) and the 
Soil Fertility Network for Maize-based Cropping Systems in Southern Africa 
(Soil Fert Net)] and many attempts are currently being made to improve commu- 
nication with NGOs and farmers. 

Targeting Solutions 

Even within narrow geographic regions a large variability exists in soil fer- 
tility such that even for adjacent fields the most suita 


ers might differ. A soil fertility replenishment strategy may best be directed 
specifically to address the rehabilitation of unproductive fields, as these can be 
left for several years under planted fallows or legume-based leys, and as the 
incremental response to addition of mineral fertilizers or high-quality organic 
residues may be greater in slightly less degraded lands (Fig. 7-5a and 7-5c). 
Where pressure for crop land is intense, interventions to improve soil fertility 
may be best targeted to the fields where the farmer has little potential yield to lose 
(Versteeg & Koudokpon, 1992). It is fairly common to see unproductive fields 
planted even though the investment of seed is scarcely warranted, presumably to 
visibly maintain ownership or for other reasons. If crops are already failing to 
yield on certain fields, then there is little loss to the farmer to break the cycle and 
use a legume cover crop or shrub for one or more seasons to boost productivity 
or rehabilitate the land for cropping. This approach has a potential disadvantage 
in that if the land is severely degraded, the soil fertility investment required to 
restore productivity is likely to be much greater and there is a risk that even the 
most robust legumes may not grow. 

Often only one legume has been recommended for soil fertility improve- 
ment when many other species or accessions of those species tested might be 
more suitable. In many African countries researchers are still investigating the 
same species tested as early as the 1920s, yet in the genus Crotalaria alone there 
are more than 480 species with 85 subspecies (Polhill, 1982). While it is possible 
that the best legumes were already identified, recent major efforts at germplasm 
collection and testing for acid soils, for example at CIAT (1993) has led to the 
identification of many new species and accessions. From knowledge of their 
adaptation and performance elsewhere, a list has been developed of forage 
legumes that would be worthwhile testing immediately in various environments 
in Africa (Thomas & Sumberg, 1995), many of which are useful as green 
manures. Improved access to information on cover crops through the use of data- 
bases (e.g., Weber et al, 1997) may assist in identification of potentially useful 

If a soil fertility replenishment strategy is envisaged in which farmers are 
offered external assistance, then such interventions could be targeted for poorly 
productive fields and the farmers recompensed for the loss of yield during the 
restorative period. Whether such an approach will survive the period during 
which assistance is offered is highly questionable considering past experiences of 
various projects, but this probably depends on how robust and successful the 
intervention is in restoring soil fertility. Even if the soil fertility is restored, care- 
ful management to conserve nutrients and continued inputs of N are necessary to 
sustain production at acceptable yields. 

The potential interventions that will most quickly lead to the build up of the 
short- to medium-term N capital of the soil are: planted legume fallows, legume- 
based ley systems, and animal manures. When land pressure dictates that fields 
cannot be rested, other means of replenishing the N supply are needed. For the 
maintenance of soil fertility under continuous cropping, repeated inputs of N 
must be assured to maintain productivity through legume rotations, intercropping 
or relay cropping with green manures, animal manures, and mineral fertilizers. 


Our knowledge of the importance of soil textural properties in the ability to 
store SOM, and hence N capital, should guide our targeting of these different 
approaches: sandy soils are going to require more frequent additions of high- 
quality organic material or mineral N fertilizers or both to sustain yields. Soils 
with a greater clay content and better structure might better be managed using 
periodic boosts to productivity such as short-term planted fallows due to their bet- 
ter ability to protect SOM, again together with some mineral fertilizer inputs and 
well-planned tillage operations. 

The urgent problem of poor crop productivity due to depleted soil fertility 
in many parts of Africa demands that ideological stances that favor the sole use 
of organic resources or the sole use of mineral fertilizers are cast aside. Supplies 
of even poor-quality organic resources for soil amendment are limited, and min- 
eral N fertilizers are undoubtedly required to ensure food production. The inte- 
grated efforts of all concerned are required to find attractive solutions to the prob- 
lem of building the soil N capital for future generations of African farmers. 


We are grateful to all those who helped by providing unpublished reports 
and references and by commenting on an earlier draft, in particular John Lynam, 
Meine van Noordwijk, and Stephen Waddington, and to the Rockefeller 
Foundation for funding much of our soil fertility research in Africa. 


igbi la \ ' mil A. A Fa em i 1971 Preliminar\ trial on th int ro ping of maize with differ- 
ent tropical legumes in western Nigeria. J. Agric. Sci. (Cambridge) 77:219-225. 

Agboola. A. A., anil A. A. Fayemi. 1972a. Effect of soil management on corn yield and sod nutrients 
in if- i i : t ii i n i hi i i , \ _ o i : ,4:u4l-6-W 

Agboola, A.A., and A.A. Fayemi. 1972b. Fixation and excretion of nitrogen by tropical legumes. 
Agron. J. 64:409-412. 

Amato. VI.. and J.N. Ladd. 1992. Decomposition of l4 C labelled glucose and legume material in soils: 
' in in In in in. ill iccumnlati no oi mil i tlue-t nil microbial biomass-C. Soil 
Biol. Biochem. 24:455 A 164. 

Amijee. F.. and K.E. Ciller. 1998. Environmental constraints to nodulation and nitrogen fixation of 
Phaseolus vulgaris L. u; Tanzania: I. A survey of soil fertility and root nodulation. Afr. J. Crop 
Sci. (in press). 

Ayed. 1 I ' ill Hi u u buti i nil n in prof i I fSndan < zira Vertisols with 

pecial referen to fixed ai inn ni un I . pi I il i _ . -61 

Bache, B.W., and R.G. Heathc< I I 9. Long-term el I if 1 itdizeis and manure on soil and 
leaves of cotton in Nigeria. Exp. Agric. 5:241-247. 

Balasubramaman, V., and L.A. Nnadi. 1980. Crop residue management and soil productivity in savan- 
na areas of Nigeria, p. 106-120. In Organic recycling in Africa. FAO Soils Bull. 43. FAO, 

Bartholomew, P.W., R. Ly, D. Kone, S.F. Traore, and K. Sissoko. 1992. Forage production for small- 

i I i i i i I i ] i 11 

Bationo, A., B.J. Ndungouru, B.R. Ntare, C.B. Christianson, and A.U. Mokwunye. 1991. Fertilizer 

in ni i in ni n ii or legume-b i i in i m in tin " I hn an semi-arid trop- 

ics, p. 213-226. //; C. Johansen et al. led.) Phosphorus nutrition ot grain legumes In die semi- 
arid tropics. Int. Crops Res. Inst, fir the Semi-Arid Tropics. Patancheru. Andhra Pradesh. 


Banei j nd I M i i I .1 ioi I carbon, nitro n. and bulk clensii > mparisons in two cropland 

tillage system i\ i i n in 1 n i nd i! Sue Am. J. 45:1 166-1 170. 

Bekunda, M.A., A. Bationo, and H Ssali i l > il fertility mana hi nl m Atuca A review of 

1 I i I I 11 II I I > ' I I II il I ' I I I I I I III I I I I Il 

Africa. SSSA Spec. Publ. 51. SSSA, Madison, WI (this publication). 
Belsky, A.J., S.M. Mwonga, and J.M. Duxbury. 1993. Effects of widely spaced trees and livestock 

grazing on undei I i pica I savann n t -> 1 

Birch. H.F. 1958. The effect of soil drvin« on humus decomposition and nitrogen availability. Plant 

Soil 10:9-31. 
Birch, H.F., and M.T. Friend. 1956. Humus decomposition in East African soils. Nature (London) 

Boudot, J.P., A.B.H. Brahim, and T. Chone. 1988. Dependence of carbon and nitrogen mineralization 

rales upon amorphous metallic constituents and allophanes in highland soils Geoderma 

Breman, H., and J.J. Kessler. 1995. Woody plants in agro-ecosystems of semi-arid regions. Springer- 

Verlag, Berlin. 
Brouwer. J., am I.M. Pi II 1 ' d i fnutri nl ii in n mi ipplication experi- 

ment in Niger, p. 435-452. In J.M. Powell et al. (ed.) Livestock and sustainable nutrient 
ling in mix II \ 1 1 [ Int. Livestock 

Ctr.-Africa, Addis Ababa, Ethiopia. 
Brown. P. 1958. The results of some short term rotation experiments in Wasaland. Rhodesia Auric. 

J. 55:626-633. 
Buckles, D. 1995. Velvetbean in the farming systems of Atlantic Honduras, p. 426-434. In H. Cook 
f 1 i! i i in i i i II Iture. Wye Coll i i 

Buresh, R.J., and G. Tian. 1997. Soil improvement by trees in sub-Saharan Africa. Agrofor. Syst. (in 

Cadisch, G., and K.E. Giller. 1997. Driven by nature: Plant litter quality and decomposition. CAB Int., 

Wallingford, England. 
Cadisch, G, R. Sylvester-Bradley, and J. Nosberger. 1989. l5 N-based estimation of nitrogen fixation 

by eight tropica] forage-legumes at two levels of P:K supply. Field Crops Res. 22:181-194. 
Cahn, M.D., D.R. Bouldin, M.S. Cravo, and WT. Bowen. 1993. Cation and nitrate leaching in an 

Oxisol of the Brazilian Amazon. Agron. J. 85:334-340. 
Campbell, B.M P.G.H. Fi H. Kirclnn ml M vift. 1 i i ycling and manage- 

ment of soil fertilit) in small-scale farming systems in north-eastern Zimbabwe. J. Sustain. 

Agric. (in press). 
Castellanos, J.Z., J.J. Pena-Cabriales, and J.A. Acosta-Gallegos. 1996. b N-determined dinitrogen fix- 
Mn n f common I n ( / ilti irs nuclei n u I siri 

(Cambridge) 126:327-333. 
Christensen, B.T. 1992. Physical fractionation of soil and organic matter in primary particle size and 

density 'parate \d\. Soil Sci. 20:1-90. 
CIAT. 1993. Invenl i tropical fora i i in utained at CIAT. Centra Intemacional de 

Agricultura Tropical, Cali, Colombia. 
! i I I I i I fj ij in i ii Ii ii i ii -Midi emnonment as 

ii i i pruning man in I. Agi i. i mbri i I 

Cooke, G.W. 1986. The intercontinental transfer of plant nutrients, p. 267-287. In 13th Congress of 

the Int. Potash Inst. Reims, France. August 1986. Int. Potash Inst., Reims. 
Cornet. F.. C. Otto. C. Rinaudo. H.G. Diem, and ') . Dommergues. 1985. Nitrogen fixation by Acacia 

liolosericeu grown in field-simulating conditions. Oecol. Plant. 6:21 1-218. 
' i II I I i I mi I ri i mi in i iii^ , in i of soil organic mat- 

i II. Ox ii i ] i i n i i mi leralization in virgin anil culti i king cla I 

Soil Biol. Biochem. 4:435 A 442. 
1 i I l i \ nl i \ \ ii i ii I I I nutrient uptake. 

Exp. Agric. 10:83-90. 
Dalai. R.C. 1989. Loiiii-term effects of no-tillage, crop residue, and nitrogen application on proper- 
ties of a Vertisol. Soil Sci. Soc. Am. J. 53:1511-1515. 
Dalai, R.C, W.M. Strong, EJ. Weston, J.E. Cooper, K.J. Lehane, A.J. King, and C.J. Chicken. 1995. 
i linin productivity i rtisol at Warra. Queensl mi with i i in i n till i_ oi 

legumes: I. Organic matter status. Aust. J. Exp. Agric. 35:903-913. 
Dancette, C, and I.F. Poulain. I Influen le 1 ir les facteurs \ limatiqi i 

les vendements kk::\ cultures. Sols Afr. 13:197-239. 


Danso, A.A., and P. Morgan. 1993a. Alley cropping maize (Zca mays var. Jeka) with cassia (Cassia 
siamni) in The Gambia: Crop production and soil fertility. Agrofor. S\ st. 21:1 33- 1 46. 

Danso, A.A., and P. Morgan. 1993b. Alley cropping rice (Oryza saliva var. Baraflta) with cassia 
(Cassia siamea): Soil fertility and crop production. Agrofor. Syst. 21:147-158. 

Davis J.H.C.. and S Garcia. 1983. Competitive ability and growth habit oi inteterminate beans and 
maize for intercropping. Field Crops Res. 6:59-73. 

Davis, J.H.C., L. van Beuningen, M.V. Ortiz, and C. Pino. 1984. Effect of growth habit of beans on 
tolerance to competition from maize when intercropped. Crop Sci 24:751-755. 

Davy, E.W. 1925. Crops for green manuring and soil protection. Agric. Dep. Bull. 3. Zomba, 

de Leeuw, P.N., MA. Mohamed-Saleem, and A.M. Nyamu. 1994. Stylosanthes as a forage and fal- 
low crop. Int. Livestock Ctr. for Africa, Addis Ababa, Ethiopia. 

Dennison, E.B. 1959. The maintenance of soil fertility in the southern Guinea zone of northern 
igeri (Ti a iinti i I , \grii (Ti ini lad) I I- 1 i 

de Ridder, N., and H. van Keulen. 1990. Some aspects of the role of organic matter in sustainable 
n i i in i I i nin i in n in ' i i n mi in I | i i > I i i i i 


de Sornay, P. 1918. Green manures and manuring in the tropics. John Bale, Sons. & Danielsson, 

Djokoto, R.K., and D. Stephens. 1961. Thirty long-term fertiliser experiments under continuous crop- 
ping in Ghana: II. Soil studies in relation to the effects ol fertilizers and manures on crop 
yields. Emp. J. Exp. Agric. 29:245-258. 

Drechsel. P.. KG. Steiner. and F. Hagedorn. 1996. A rex iew on the potential of improved fallows and 
green manure in Rwanda. Agrofor. Syst. 33:109-136. 

Dudal, R., and J. Deckers. 1993. Soil organic matter in relation to soil productivity, p. 377-380. In K. 
mi ii in 111 lii ii n i i I ii in i i ii i n I i f i i i I 

agriculture. John Wiley * Sons. Chichester. England. 

Duguma, B., B.T. Kang, and D.U.U. Okali. 1988. Effect of pruning intensities of three woody legu- 
minous species grown in alley cropping with maize and cowpea on an Alfisol. Agrofor. Syst. 

Duguma. B.. and J. Tonye. 1994. Screening of multipurpose tree and shrub species for agroforestry 
in the humid lowlands of Cameroon. For. Ecol. Manage. 64:135-143. 

Dunham, K.M. 1989. Litterfall, nutrient-fall and production in an Acacia albida woodland in 
Zimbabwe. J. Trop. Ecol. 5:227-238. 
"ii! i n I 1 i ii I i i i mi 1 i n i i hi I I I i 

terns in South Africa. Fert. Res. 42:27-32. 

Dupuy, N., and B. Dreyfus. 1992. Presence of bradyrhizobia under Acacia albida. p. 145-148. In R.J. 
Vandenbeldt (ed.) Faidherbia albida in the West African semi-arid tropics. ICRISAT/ICRAF, 
Andhra Pradesh, India, and Nairobi, Kenya. 

Eaglesham, A.R.J., A. Ayanaba. V. Rangy Rao. and D.L. Eskew. 1982. Mineral N effect on cowpea 
ami soybean crops in a Nigerian soil: Ii. Amounts of N fixed and accrual to flic soil. Flam Soil 

Elliot, E.T 1986 Aii« I in an nul irb n niti " i nul pho Ii i >, in native and cultivated 
soils. Soil Sci. Soc. Am. J. 50:627-633. 

Elliott, E.T, J.A. Anderson, D.C. Coleman, and C.V. Cole. 1980. Habitable pore space and microbial 
ii i In intei iction « il . ?5 327-335. 

Ensminger, L.E., and J.E. Giesekinti. 1942. Resistance ol clav-adsorbed proteins to proteolytic 
hydrolysis. Soil Sci. 53:205-209. 

Eswaran, H., J. Kimble, T Cook, and F.H. Beinroth. 1992. Soil diversity in the tropics: Implications 
for agricultural development, p. 1-16. In R. Lai and P.A. Sanchez led.) Myths and science of 
soils of the tropics. SSSA Spec. Publ. 29. SSSA, Madison, WI. 

Feller, C, E. Fritsch, R. Poss, and C. Valentin. 1991. Effet de la texture sur la stockage et la 
dvnamique ties matieres oraanique dans quelques sols ferrusjineux et ferralitiques Cahiers 
ORSTOM Ser. Pedologique 26:25-36. 

Fernandez-Rivera, S., TO. Williams, P. Hiemaux, and J.M. Powell. 1995. Faecal excretion by rumi- 
nant and m u i 1 I li I i i hi I i i p 149-169. In 
'IP i 1 I ( i ' mi ii il urn ii I i i ixed farmii i i i 
of sub-Saharan Africa. Int. Livestock Ctr. for Africa, Addis Ababa, Ethiopia. 

1 1 I 1 I ii t n 

i > mi l i I I i n 1 i 1 ii 1 ii Ph D. diss. ETH- 

Inst. of Plant Sciences. Zurich. Switzerland. 


Franzluebbers, K., A.S.R. Juo, and A. Mann. 1994. Decomposition of cowpea and millet amendments 

to a sandy Alfisol in Niger. Plant Soil 167:255-265. 
Gethin Jones, G.H. 1942. The effect o1 i summon i vera p in building up oil fertility. East Air. 

Agric. J. 8:48-52. 
Giller, K.E., F. Amijee, S.J. Brodrick, and O.T. Edje. 1998. Environmental constraints to nodulation 

and nitrogen ti tion of P oh L. in Tan; nia 11 ''esponse to N and P fertilizers 

ami inoculation. Afr. ,1. Crop Sci. (In press). 
Giller, K.E., and G. Cadisch. 1995. Futuir b n :fit from biologii il n trogen fixation: An ecological 

approach to agriculture. Plant Soil 174:255-277. 
Giller, K.E., and J.M. Day. 1985 Nitrogen fixation in t on I i i in ance in natural and 

agricultural systems, p. 127-147. /// A.H. Fitter (eel) Ecological interactions in soil: Plants, 

microbes and annual Blackwell i a fi < \toid England. 
Giller, K.E., J.F. McDonagh, and G. Cadisch. 1994. Can biological nitrogen fixation sustain agricul- 
ture in the tropics? p. 173-191. In J.K. Syers and D.L. Rimmer (ed.) Soil science and sustain- 
able land management in the tropics. CAB Int., Wallingford, England. 
Giller. I i ( > i i li nisi 1 II ransfei i i bean to ii 

cropped maize measured using h N-enrichment ami X-isotope dilution methods. Soil Biol. 

Biochem. 23:339-346. 
Giller, K.E., and K.J. Wilson. 1991. Nitrogen fixation in tropica! cropping systems. CAB Int., 

Wallingford, England. 
Golchin, A., J.M. Oades, J.O. Skjemstad, and P. Clarke. 1994a. Soil structure and carbon cycling. 

Aust. J. Soil Res. 32:1043-1068. 
Golchin, A., J.M. Oades, J.O. Skjemstad, and P. Clarke. 1994b. Study of free and occluded particu- 

1 ite ii mi in H i n I \ ii ir i » i I nu ] n I innin elec- 

tron microscopy. Aust. J. Soil Res. 32:285-309. 
Grant, P.M. 1967. The fertility ofsandveld soil under continuous cultivation: I. The effect of manure 

and nitrogen fertilizers on the nitrogen status of the soil. Rhod. Zambia Malawi J. Agile. Res. 

Greenland, D.J. 1958. Nitrate fluctuations in tropical soils. .1. Agric. Sci. (Cambridge) 50:82-91. 
i il ii III I. Long-term croppi s inn n in I Ii | in uitri ill i r 1 1 > 

in 1 I ' Leigh and A.E. Johnston led i i u 1 1 

in agricultural and e ilogical scien < ,t' hit '> illin toid. England. 
Greenland, D.J., and P.H. Nye. 1959. Increase in the carhop, and nitrogen contents of tropical soils 

under natural fallows. J. Soil Sci. 10:284-299. 
Greenland, D.J., A. Wild, and D. Adams. 1992. Organic matter dynamics in soils from the tropics — 

from mvtl mpl reali i 1 In R. Lai and P nch i Ivt ml in 

of soils of the tropics. SSSA Spec. Publ. 29. SSSA, Madison, WI. 
ill i t I I ii n ii II i i II 

Trap. Grassl. 20:107-112. 
Haggar, J.P., E.V.J. Tanner, J.W. Beer, and D.C.L. Kass. 1993. Nitrogen dynamics of tropical agro- 
forestry and annual cropping systems. Soil Biol. Biochem. 25:1363-1378. 
Hairiah. K. 1993. Aluminum tolerance ot Miiama. a tropical leguminous cover crop. Ph.D. diss 

il 1 i i i it « I nu n ' i in u I ill II 11 

Halvin, J.K., D.E. Kissel, L.D. Maddux, M.M. Claassen, and J.H. Long. 1990. Crop rotation and 

nil iii il i ni irbon and niti i i I i Mn I 54448 A 152. 

Handayanto, E., G. Cadisch, and K.E. Giller. 1994. Nitrogen release from primings of legume 
i i tiees in lelation to quality of the pun n in I incubatioi u thod. Plant Soil 

Handayanto, E., G. Cadisch, and K.E. Giller. 1995. Decomposition and X mineralization of selected 

hedgerow tree primings in relation to their nitrogen, lignin and polyphenol contents, p. 

113-122. In H. Cook and H. Lee (ed.) Soil management in sustainable agriculture. Wye 

College Press, Wye, Kent, England. 
Handayanto, E., K.E. Giller, and G. Cadisch. 1997. Nitrogen mineralization from mixtures of legume 

tree primings of different quality and recovery of nitrogen by maize. Soil Biol. Biochem. 

Harden. H.J. 1936. Factoren. die het organische stof- en het stickstof-gehalte van tropische gronden 

h ii hen (Fact ntn Ming th n n mattei nu ill nil i i ill ill oi trot u ul I 

Landbouw 11:517-540. 
Hartemink, A.E., R.J. Buresh, B. Jama, and B.H. Janssen. 1996. Soil nitrate and water dynamics in 

sesbania fallows, weed fallows, and maize. Soil Sci. Soc. Am. J. 60:568-574. 


Hartley. K.T. I <J 3 7 . An explanation of the effect of farmyard manure in northern Nigeria. Emp. .1. Exp. 
Agric. 5:254-263. 

Hi nl I lW5a.De mposition rate const ml f si nil density fractioi > 1 soil oigamc matter. 
Soil Sci. Soc. Am. J. 59:1631-1635. 

Hassink. I. IW5b. Organic matter dynamics and N mineralization in grassland soils. Ph.D. diss 
mcultural I ni\ ' j nin : n th ih 1 1 md 

Hassink. .1. IW6. Preservation of plant residues in soils differing in unsaturated protective capacity. 
Soil Sci. Soc. Am. J. 60:487-491. 

Haynes. R.I. I9S6. The decomposition process: Mineralization, immobilization, humus formation and 
degradation, p. 52-109. /// R.J. Haynes led.) Mineral nitrogen in the plant-soil system. 
Academic Press, Madison, WI. 

Heal, O.W., J.W. Anderson, and M.J. Swift. 1997. Plant litter quality and decomposition: An histori- 
cal overview, p. 3-45. In G. Cadisch and K.E. Giller (ed.) Driven by nature: Plant litter qual- 
ity and decomposition. CAB Int., Wallingford, England. 

Heisey, P.W., and W. Mwangi. 1996. Fertilizer use and maize production in sub-Saharan Africa. Int. 
I i i Vheat Ini] Cti 1 « I 

Herrmann, L. 1996. Staubdeposition auf boden West-Afrikas Eigenschaften und herkunftsgebiete der 
staube und ihr einflul ml >od n unci i nd I igen i itten. Ph.D. diss. Univ. of Hohenheim, 
Hohenheim, Germany. 

Itimti. O A. 1997. Nitrogen dynamics and mot distribution of Glirickliu xepiinn and Senna .speclabilh 
n i i I lil iii i i ii 1 I i 1 I i I L'niv. 

of London, Wye, Kent, England. 

Jaiyebo. E.G.. and A.W. Moore. 1964. Soil fertility and nutrient storage in different soil-vegetation 
systems in a tropical rainforest environment. Trap. Agile (Trinidad) 41 : 129-1 39. 
na, B., R.J. Buresh, J.K. Ndufa, and K.D. Shepherd. 1998. Vertical distribution of roots and soil 
nitrate as attei ted by h pi ies and phosphorus. Soil Sci. Soc. Am. J. (in press). 

Jenkinson. D.S. 19/7. Studies on the decomposition of plant material in soil: V. The effects of plant 
cover and soil type on the loss of carbon from 'C-labelled ryegrass decomposing under field 
conditions. J. Soil Sci. 28:424-434. 
enkinson, D.S., D.E. Adams, and A. Wild. 1991. Model estimates of CO, emissions from soil in 

response to global warming. Nature (London: 351:304-306. 
enkinson. D.S.. and A. Ayanaba. 1977. Decomposition of carbon- 1 4 labelled plant material under 
tropical conditions. Soil Sci. Soc. Am. J. 41:912-915. 

John, P.S., R.K. Pandey, R.J. Buresh, and R. Prasad. 1992. Nitrogen contribution of cowpea green 
manure and residue to upland rice. Plant Soil 142:53-61. 
i, M.J. 1973. The organic matter of the savanna soils of West Africa. J. Soil Sci. 24:42-53. 
mi l 1 i ill i f tin mi i n rtili i md im n pH hi I u-1 ngeable cation con- 
tent at different depths in cropped soils at two sites in the Nigerian savanna. Trop. Agric. 
(Trinidad) 53:243-254. 
s, M.J., and A.R. Bromfield. 1970. Nitron n in IL i lin all a ' im an iseri Nature (London) 

>, M.J., and A. Wild. 1975. Soils of the West African savanna. Tile maintenance and improve- 
ment of their fertility. Tech. Commun. Commonwealth Bureau of Soils. Harpenden, England. 

Kannegieter, A. 1968. Zero cultivation and other methods of reclaiming Pneraria fallowed land for 
food crop cultivation in the forest zone of Ghana. Trop. Agric. J. (Ceylon) 123:51-73. 

Kapkiyai, J. 1996. Dynamics of soil organic carbon, nitrogen, and microbial biomass in a long-term 
experiment as affected by inorganic and organic fertilizers. M.S. thesis. Dep. of Soil Science, 
University c i iii in in bi I in 
hi inch I V in to in I i in nil i 199 Crop yield md ul i J itiht\ trends in cur- 

rent maize-with-groundiuit cropping patterns on smallholder farms in Zimbabwe, p. 54-64. In 
S.R. Waddington (eel.) Research results and network outputs in 1994 and 1995. Proc. of the 
2nd Meeting of the soil fertility working group, Kadoma, Zimbabwe. July 1995. CIMMYT, 
Harare, Zimbabwe. 

Klaij, M.C., C. Renard, and K.C. Reddy. 1994. Low input technologj options tor millet-based crop- 
ping systems in the Sahel. Exp. Agric. 30:77-82. 

Kouame, C.N., J.M. Powell, C.A. Renard, and K.H. Quesenberry. 1993. Plant yields and fodder qual- 
ity related characteristics of millet-stylo intercopping systems in the Sahel. Agron. J. 

Kumar Rao, J.V.D.K., and P.J. Dart. 1987. Nodulation. nitrogen fixation and nitrogen uptake in 
pigeonpea tCujantts cajtin iL.i Millsp.) of different matin it) groups. Plant Soil 99:255-266. 


Kumar Rao, J.V.D.K., PJ. Dart, and P.V.S.S. Sastry. 1983. Residual effect of pigeonpea (Cajanus 

cajan) on yield and nitrogen response of maize. Exp. Agric. 19:131-141. 
Kumar Rao, J.V.D.K., J.A. Thompson, P.V.S.S. Sastry, K.E. Giller, and J.M. Day. 1987. Measurement 

i ii ii in ii 1 I i v ii i i , t (L.) Mil Isp.J usin I ib II l fer- 

tilizer. Plant Soil 101:107-113. 

Kwaye, P.A. 1980. The etrea if in i 1 lot un t i - n i 1 niunent (N-P-K) content and crop 
yield in northeast e mini >i h n p -2 .. / II n oforgani material fertiliz i 

in Africa. FAO, Rome. 

Kwaye, P.A., E.A. Dennis, and A.E. Asmah. 1995. Management of a continuously cropped forest soil 
through fertilizer use. Exp. Agric. 31:325-331. 

Kwesiga F and P < ] 4. Tl feet of short rotation i planted fallows on maize 

yield. For. Ecol. Manage. 64: 199-208. 

Ladd, J.N., M. M. Amato, P.R. Grace, and J.A. van Veen. 1995. Simulation of 'X turnover through 
the microbial biomass in soils incubated with X-labelled plant residues. Soil Biol. Biochem. 

I I I llii ID I i At ti ill in soils: III. 

Residual organic and microbial bioinass-C and biomass-N from isotope-labelled legume mate- 
ria.! and soil organic-matter, decomposing under field conditions. Aust. J. Soil Res. 

'.: Carbon and nitrogen metabolism. 

Ladha, J.K., M.B. Peoples, D.P. Garrity, V.T. Capuno, and P.J. Dart. 1993. Estimating dinitrogen fix- 

ii ii row i i i i i i i i Hi n- 1 5 natural al urn n meth i 

Am. J. 57:732-737. 

Lai. R. 1974 ■ -Mil effects on soil propertii an in i iZea /hows L.) j luction in western 
Africa. Plant Soil 40:321-331. 

Lai. R. 19 i rill : it i in oil proj it tin I i dif1 rent ps in i in Nigeria. Soil Sci. 

Soc. Am. J. 40:762-768. 

Le Mare, PH. 1972. A long term experiment on soil fertilitx and cotton yield in Tanzania. Exp. Agric. 

Ledgard, S.J., and K.E. Giller. 1995. Atmospheric Nj-fixatioi 
443-486. In P. Bacon (ed.) Nitrogen fertilization and t 

MacColl, D. 1989. Studies on maize (Zea mays) at Bunda, Malawi: II. Yield in short rotations with 
legumes. Exp. Agric. 25:367-374. 
I I n i I ill i 1 i ii I i I i i iction of soil organ- 

ic matter: Three size-densit\ fractionation methods for tracing the fate of homogeneously X 
labelled plant materials in soil. Soil Biol. Biochem. 28:89-99. 

Mandimba. G.R. 1995. Contribution of nodulated legumes on the growth ot'Zeo maw, L. under vari- 
ous cropping systems. Symbiosis 19:213-222. 

Matthews, R.B., S. Lungu, J. Volk, ST. Holden, and K. Solberg. 1992. The potential of alley crop- 
in III ill ii 1 in llti i t I 1 ill Ii II i 111! II 1 l u II. M , 

duction. Agrofor. Syst. 17:241-261. 

McDonagh, J.F B I<>on m ^ Limpinuntai null i Gill I 1 ". I timat of the residual nitro- 
gen benefit of groundnut to maize in Northeast Thailand. Plant Soil 154:267-277. 

McDonagh, J.F., B. Toomsan, V. Limpinuntana, and K.E. Giller. 1995 Grain legumes and green 
manures as pre-rice < rops in Northeast Thailand: I. Legume nitrogen fixation, production and 
residual nitrogen benefits to rice. Plant Soil 177:127-136. 
I ee I n . O.I i i B ij ii! i 1 1 i i I ( ductivity improve- 

ment. Food and Agric. Organ., Rome. 

Mekonnen. K.. R..I Buresh. and B. Jama 1997. Root and inorganic nitrogen distributions in sesbania 
fallow, natural fallow and maize fields. Plant Soil 188:319-327. 

Morris, R.A., R.E. Furoc, and M.A. Dizon. 1986. Rice responses to a short-duration green manure: 
II. N recovery and utilization. Agron. J. 78:413-416. 

Motavalli, P.P., C.A. Palm, E.T Elliott, S.D. Frey, and PC. Smithson. 1995. Nitrogen mineralization 

in humid tropical rest i lii I text nd i i i fraction il 

Soc. Am. J. 59:1168-1175. 
Ipepereki si ' i nd K.E. Giller. 199 beans in sm dli lder 1 rutin i in in 

Zimbabwe. CIMMYT, Harare, Zimbabwe. 

Mueller-Harvey, I., A.S.R. Juo, and A. Wild. 1985. Soil oi 


Mulongoy, K., and I.O. Akobundu. 1985. Nitrogen uptake of maize in live mulch systems, p. 

1 ' I- 1 ' M ii I in der Heid In n man ment in fa min i in 

in humid and suhlniinid tropics. IITA. Ibadan. Nigeria. 
Murwira. H.K 1995. Ammonia losses from Zimbabwean cattle manure before and after incorpora- 
tion into soil. Trop. Agric. (Trinidad) 72:269-276. 
Murwira. H.. and H. Kircliiiiann. 1993. Carbon and nitrogen mineralization of cattle manures, sub- 
I i<i i hi in i hi i hi in hi in n in lish soils, p. 1 89-1 ' i \ ml n 

and R Merckx (ed.i Soil organic matter dynamics am! sustainability of tropical agriculture. 
John Wiley & Sous. Chichester. England. 
Iwiin II 111 hi' 'mi l''r T" ii f i n i I of six multipurpose 

I 111 1 

Nambiar, P.T.C., M.R. Rao, M.S. Reddy, C.N. Floyd, RJ. Dart, and R.W. Willey. 1983. Effect of inter- 
cropping on nodulation and N-fixation b\ groundnut. Exp. Agric. 19:79-86. 
Nommik, H., and K. Vahtras. 1982. Retention and fixation of ammonium and ammonia in soils, p. 
123-172. In F.J. Stevenson (ed.i Nitrotien in atiricultural soils. Agron. Monogr. 22. ASA, 
CSSA, and SSSA, Madison, WI. 

t I 1 II I I I | I II 

n in i i 1 i ni I i f the Sudai i n i t Africa. ] 

li I i in i i ii ni i in iii ii uliii i ill Mi i ii 

/one [in Crops Res. Inst, foi the Semi-Arid Tropics. Patancheru. Andhra Pradesh, India. 
Nyborg, M., E.D. Solberg, I. R.C., S.S. Malhi, and M. Molina- Ayala. 1995. Influence of long-term 

till I e tl 1\\ II I I ll II I 1 \ I | I Liptal I ill 1 III e ll I |l| [ 


Nye, P.H., and DJ. Greenland. 1960. The soil under shifting cultivation. Commonwealth Agric. 
Bureaux, Harpenden, England. 

Oades, J.M., and A.G. Waters. 1991. Aggregate hierarchy in soils. Aust. J. Soil Res. 29:815-828. 

Oades, J.M., A.G. Waters, A.M. Vasallo, M.A. Wilson, and G.P. Jones. 1988. Influence of manage- 
inent on the composition of organic mallei in a red-brown earth as shown b\ C nuclear mag- 
netic resonance. Aust. J. Soil Res. 26:289-299. 

('fori, f I | i ii I, I I i i 1 -fixation ami niti i.en economy of a 

maize/cowpea mtei i mi n 1 i I ition methods. Plant Soil 102:149-160. 

Ofori, F., and W.R. Stern. 1987. Ceieal I ume int i ( i. terns Adv. Agron. 41:41-90. 

> > 1 1 i i i n i r i i i i i 1 1 i 1 1 i pi > i 1 1 ii 

iZea maw) and beans (Plia\eolus vulgaris) at Moroiioro. Tanzania. For. Ecol. Manaae. 
Palm. C.A.. R..I.K. Myers, and S.M. Nandwa. 1997. Combined use of organic and inorganic nutrient 

i iii rtili in n ml replenishment, p. 19 I I ; Ii i i I 

Replenishing soil fertility in Africa. SSSA Spec. Publ. 51. SSSA, Madison, WI (this publica- 

Parton, W.J., D.S. Schimel, C.V. Cole, and D.S. Ojima. 1987. Analysis of factors controlling soil 
organic matter le\ el in ( it Plaii i hi Soil Sci. Soc. Am. J. 51:1173-1179. 

Patel, L.B., B.S. Sidhu, and V. Beri. 1996. Symbiotic efficiency of Sesbania rostrata and S. cannabi- 
na as affected by agronomic practices Biol. Fert. Soils 21:149-151. 

Peoples, M.B., F.J. Bergersen, G.L. Turner, C. Sampet, B. Rerkasem. A. Bhromsiri, D.P. Nurhayati, 

A.W. Faizah. M.N. Stidin. M. Norhavati. ami D.F. Herridge. IVOR Rse of natural enrichment 

of''N in plant available soil N for the measurement of symbiotic N 2 fixation, p. 117-130. In 

II i i |l i i I i i i i i i i I i i f ^ )iii 

Peoples, M.B., and E T i II I >9 Biological nitro i-fi ii — in tin ni [ tatioi ml 

actual contributions to agriculture. Plant Soil 141:1 3-39. 
Peoples, M.B., B. Palmer, D.M. Liiley, L.M. Due, and D.F. Herridge. 1996. Application of l5 N and 

\ ill tireide methods foi hit: fixation ol th inn I um »!i ill prui d I i 

forage. Plant Soil 182:125-137. 
Pemet, R. 1952. L'humus forestier el -ion evolution apres destruction du couverte vegetale. Mem. 

Inst. Sci. Madagascar Ser. D 4:20 1 -2 1 9. 
Pit-hot. J ii (in I Ii ill I r I i ii n nl pul I in F i ii I i ii 

dans les sols sableux dunaires. Agron. Trop. 24:995-1005. 
Pichot, J., M.P. Sedogo, J.F. Poulain, and J. Arrivets. 1981. Evolution de la fertilite d'un sol ferrug- 

ii ii 1 1 il us fulfill d i nin inn I i ij i ^ i liop 36:122-133. 


Pieri, C.J.M.G. 1992. Fertility of soils: A future for farming in the West African savannah. Spnnser- 
Verlag, Berlin. 

Pinck, L.A., R.S. Dyal, and F.E. Allison. 1964. Protein-montmorillonite complexes, their preparation 
and the effects of soil microorganisms on their decomposition. Soil Sci. 78:109-118. 

Pleysier, J.L., and A.S.R. Juo. 1981. Leaching of fertilizer ions in an Ultisol from the high rainfall 
tropics — leaching through undisturbed soil columns. Soil Sci. Soc. Am. .1. 45:7n4-760. 

Polhill, R.M. 1982. Crotalaria in Africa and Madagascar. A.A. Balkema, Rotterdam, the Netherlands. 

Powell, J.M. 1986. Manure for cropping: A case study from central Nigeria. Exp. Agric. 22:15-24. 
*o\ 111 ill 'bund leem. 19 iti en in ill pho u tninsfei in i [ li 

stock system in West Africa. Agric. Syst. 25:261-277. 

Powell. .1 VI.. and TO. Williams. 1993. Livestock, nutrient cycling and sustainable agriculture in the 
West African Sahel. IIED, London. 

Powlson. D.S. 1980. The effects ot grinding on microbial anil non-niicrobial organic matter in soil. J. 
Soil Sci. 31:77-85. 

Probert, M.E., J.R. Okalebo, and R.K. Jones. 1995. The use of manure on smallholders' farms in 
semi-arid eastern Kenya. Exp. Agric. 31:371-381. 

Prudencio, C.F. 1993. Ring management of soils and crops in the West African semi-arid tropics: The 
case of the mossi farming system in Burkina Faso Ague Ecosyst Environ. 47:237-204 

• )iiil] n I I ind [ OH ill I » i iu I ultui n fumui ii ii i il L'na pei t des relations 
agriculture-elevage dans le nord de la Haute-Volta. Agron. Trop. 38:206-212. 

Rattray, A.G.H., and B.S. Ellis. 1952. Maize and green manuring in Southern Rhodesia. Rhod. Agric. 
J. 49:188-197. 

Reddy, K.C., PL. Visser, M.C. Klaij, and C. Renard. 1994. The effects of sole and traditional inter- 
cropping of millet and cowpea on soil and nop productivity. Exp. Agric. 30:83-88. 
I I II I I I I l ep Int ike i wth 

Mi III phenol i i I nim. Feed I ' i 

Reynolds, S.G. 1982. Contributions to yield, nitrogen fixation and transfer by local and exotic 
legumes in pical gra I in mi u ii n Samoa. Tro] i I 

Robinson CA R i < n mil haffai lei I99i i n tern and niti n effeel n 

Mollisol organic carbon. Soil Sci. Soc. Am. J. 60:264-269. 

Robinson, J.B.D. 1957. The critical relationship between soil moisture content in the region of wilt- 
ing point and the mineralization of natural soil nitrogen. J. Agric. Sci. (Cambridge) 

Rodel, M.G.W., J.D.H. Hopley, and J.N. Boultwood. 1980. Effects of applied nitrogen, kraal compost 
ami maize stovei on the yields of maize crown in L 101 gi mite 'I iJmbabw Am I 

Rovira, A.D., 

Sanchez. P. A. 1970. Properties and management of soils in the tropics. John Wiley. New York. 

Sanchez, P.A. 1995. The science ofagroforestry. Agrofor. Syst. 30:5-55. 

Sanchez, P.A., K.D. Shepherd, M.J. Soule, F.M. Place, A.U. Mokwunye, R.J. Buresh, F.R. Kwesiga, 
A.-M.N. Izac, C.G. Ndiritu, and PL. Woomer. 1997. Soil fertility replenishment in Africa: An 
ii Hi nl m i i n I i i i ii ] n[ | I — 16. /// R.J. Buresh et al I I In bin il fer- 

tility in Africa. SSSA Spec. Pub. 51. SSSA, Madison, WI (this publication). 

Sandford, S.G. 1989. Crop residue/livestock interactions, p. 169-182. In Soil, crop, and water man- 
in 1 I I I II 11 nil I 111 1 II II II II lilt 1 11 I I 

the Semi-Arid Tropics. Patancheru. Andhra Pradesh India 
Sanginga, N., B. Ibewiro. P. Houngnandan. B. Vaniauwe. J. A. Okogun. l.O. Akobundu. and M 

Versteeg. 1996. Evaluation of symbiotic properties and nitrogen contribution of muciina to 

maize grown in the derived savanna of West Africa. Plant Soil 179:119-129. 
Sanginga. N.. K. Mulongoy. and A. Ayanaba. 1986. Inoculation of Lciiciiena leitcm cpliala (Lam.) de 

Wit ii i i i n i i in i i i i i i H ii in i i I 

Hortic. 3:347-352. 
Sanginga, N., K. Mulongoy. and A. Ayanaba 1989. Nitrogen fixation or field-inoculated Leiicaenti 

leacocepluila (Lam. I de Wit estimated by the ,r, N and the difference methods. Plant Soil 

Schulze. E.D.. G. Gebauer. H. Ziegler. uvul O.L. Lange. 1991. Estimates of nitrogen fixation by trees 

on an aridity gradient in Namibia. Oecologia 8:431-459. 
Seneviratne. R anc \. Wild. 19) Eft t of mild di in on tb minei liz ion of soil nitrogen Plant 

Soil 84:175-179. 


Seyfried, M.S., and P.S. Rao. 1991. Nutrient leaching loss from two contrasting cropping systems in 

the humid tropics. Trop. Agric. (Trinidad) 68:9-18. 
Shumba, E.M., H.H. Dhilwayo, and O.Z. Mukoko. 1990. The potential of maize-cowpea intercrop- 
ping in low rainfall areas of Zimbabwe. Zimbabwe J. Agric. Res. 28:33-38. 
Sibancl. P. I l >74. Evolution des earacteres et tie la fertilite d'un so! roime lie Casamance. Aaron. Trop. 

Simpson, J.R. 1961. The effects of several agricultural treatments on tin- nitroiien status of a red earth 

in Uganda. East Afr. Agric. For. J. 26:158-163. 
Sisworo, W.H., M.M. Mitrosuhardjo, H. Rasjid, and R.J.K. Myers. 1990. The relative roles of N fix- 
ii n 1 i i i i i u ii ii n 1 1 ii i i i iln I i in i n in i i n il 

i i I hi I i i | m tern. PI m il I I 

Skjemstad, J.O., P. Clarke, A. Golchin, and J.M. Oades. 1997. Characterization of soil organic matter 

by' 13 CNMR j. i [i 71. ///G.i icli Ii n I nl i > i Driven by nature: 

Plant litter quality and decomposition. CAB Int., Wallingford. England. 
Smaling, E.M.A.. and O. 1 I l! lecision-support model I noni ng nutrient balances 

under agricultural land use (NUTMON). Geoderma 60:235-256. 
Smaling, E.M.A., S.M. Nandwa, and B.H. Janssen. Soil fertility in Africa is at stake, p. 47-61. fa R.J. 

Buresh et al. (ed.) Replenishing soil fertility in Africa. SSSA Spec. Publ. 51. SSSA, Madison, 

WI (this publication). 
Smaling, E.M \ I I i ' i i ' Wintlmeij 1 Calculath n nutrient balances in 

Africa at different scales: II. District scale. Fert. Res. 35:237-250. 
Smithson, J.B., O.T. Edje, and K.E. Giller. 1993. Diagnosis and correction of soil nutrient problems 

of common bean (Pha.sailti.s vulgaris) in the Usambara Mountains of Tanzania. J. Auric. Sci. 

(Cambridge) 120:233-240. 
Somda, Z.C., J.M. Powell, S. Fernandez-Rivera, and J. Reed. 1995. Feed factors affecting nutrient 

excretion by ruminants and the fate of nutrients when applied to soil. p. 227-246. In J.M. 
Ii 1 i I i\ i i nil sustainabl u i i i i lin u 1 farmii i i i i i 

Saharan Africa. Int. Livestock Ctr. foi Africa. Addis Ababa, Ethiopia. 
Sarensen. L.H. 19 1 biliz tioi ofnewh i on o imini m tabolil n ij| !>\ I,., mil ral 

Soil Sci. 114:5-11. 
Soumare. A.. J. J.R. Groot. D. Kone. and S. Radersma. 1993. Structure spatiale du systeme racinaire 

de deux arbes du Sahel: Acacia seyal el Sclerocarya birrea. Inst. il'Economie Rurale, 

Bamoko, Mali and AB-DLO. Wageningen. the Netherlands. 
Ii. H.. ai i 1 i ii u ! I i i lu i i i ill l ii I u i n 

and yield ol gi un 1 j mi 1 Cowj till u East Af nc For. J. 45:247-254. 
Ii. H.. and S.O. I I ti n level and ilti i n nodul n u liniti n i tion 

and yield of grain legumes: II. Common bean cuitivars. East Afr. Agric. For. J. 45:277-283. 
Ssali, H., and S.O. Keya. 1986. The effect if ph I hi ii niti nl rtiliz le\el on nodulation, 

i I 1 I 1 1 I I I v til II 

Stoorvogel. J. J.. E.M. A. Smalmg. anil B.H. Janssen. 199.3. Calculating soil nutrient balances in Africa 
t different lit II ipra-national I rt. Rt 
ii u\ u nil S. Cliotchungman i I uantiti i i i I i i 

iirain letuunes on foilowinti maize, and X and P status of soil as indicated bv isotopes. Plant 
Soil 93:249-258. 

h I II III I i i i 1 n i i I n il 

humid zone of Nigeria. J. \gric. Sci. (Cambridge) 123:55-60. 
Tarawali, S.A., and M. Peters. 1996. The potential contribution of selected forage legume pastures to 
cereal production in ci op I I i rming systems. J. Agric. Sci. (Cambridge) 127:175-182. 

I 1 I f i ti I propei I | il 1 lei Amsterdam, the 

Theng, B.K.G., G.J. Churchman, and R.H. Newman. 1986. The occurrence of interlayer clay-organ- 
ic complexes in two New Zealand soils. Soil Sci. 142:262-266. 

Theng, B.K.G, K.R. Tate, and P. Soilings. 1989. Constituents oi organic matter in temperate and trop- 
ical soils, p. 5-32. In D.C. Coleman, M. Oades andG. Uehara (ed.) Dynamics of soil organic 
in i i i ii i i i in 1 i i ll \ n ' 1 n lull 

(j in I ii I 1 i ii 1 revi i luation and i I pical forage legumes 

in sub-Saharan Africa. Agric. Ecosyst. Environ. 54:151-163. 

I I ' I i ill ill h ' hm ti t[ I i i i i i ith h hi ll n 

trusting compositions on maize growth ami nutrients accumulation. Plant Soil 153:179-187. 
1 1 n 1 1 nid I.W.B i ii i < Particle-siz fraction mil tl n u in indies of soil organic 
matter: i! Cultivation effects on organic matter composition in size fractions. Soil Sci. Soc. 
Am. J. 47:509-514. 


i] : iiil soils. J. Soil Sci. 

Toomsan, B., J.F. McDonagh. V. Linipinuntana. and K.E. Ciller. 1995. Nitrogen fixation by ground- 
nut and soybean and residual nitrogen benefits to nee in farmers' fir-Ids in Northeast Thailand. 
Plant Soil 175:45-56. 

Vandenbeldt, R.J. 1992. Faidherbia albida in the West African semi-arid tropics. ICRISAT/ICRAF, 
Andhra Pradesh. India, and Nairobi. Kenya. 

van der Kruijs, A.C.B.M., M.T.F. Wong, A.S.R. Juo, and A. Wild. 1988. Recovery of - 5 N-labelled fer- 
tilizer in crop di tin m watet n oil usin mini ilith I- ini tei in si ttth i f Jigeria 1 Soil 
Sci. 39:483-492. 

Van Kessel, C, and J.P. Roskoski. 1988. Row spacing effects on N ; fixation, N-yield and soil N 
uptake of intercropped cowpea and maize. Plant Soil 1 I 1:17-23. 

\ nl ir> B 1996. Residu qtialit) . mpo ition nn id i , nn in n , hnamics under sub- 
i i i i ii pi I liti Ph. I i i hi h 1 in it 1 i ii ' ii n 

Vanlauwe, B., M.J. Swift, and R. Merckx. 1996. Soil litter dynamics and N use in a leucaena 
( L * i 11 i | hwestern Nigeria. Soil 

Biol. Biochem. 28:739-749. 

van Noordwijk Ml ting depth in eroppin tents ii th humid tt p in relation to nutri- 

ent use iii i n ■ i I i- 1 1 ' el in tier Heid led ntrient man : ment for food crop 
in in i in in i i 1 i ii hi i in In i ill nI i Haren. th ' i l 1 1 i i 

van Noordwijk, M. 1996. Mulch and shade model for optimum alley-cropping design depending on 
soil fertility, p. 51-72. /;/ I >ie_ .m I ^ Huxl it I ii <i pin ■! i ions: A physiolog- 
ical approach. CAB Int., Wallingford, England. 

van Noordwijk, M., C. Cerri, PL. Woomer. K. Nugroho, and M. Bemoux. 1997. Soil carbon dynam- 
ics in the humid tropical forest zone. Geoderma 79:187-225. 

van Noordwijk, M., and P. Purnomisidi. 1992. Litter fall in cassava-based cropping systems in 
Lampiing. Agrivita 15:29-33. 

van Veen, J.A., J.N. Ladd, and M. Amato. 1985. Turnover or carbon and nitrogen through the micro- 
bial biomass in a sandy loam and a clay soil incubated with [Xdojglucose and 
[ 15 N](NH4)S0 4 under different moisture regimes. Soil Biol. Biochem. 17:747-756. 

van Veen, J.A., J.N. Ladd, J.K. Martin, and M. Amato. 1987. Tin nee. er of carbon, nitrogen and phos- 

h in mi 'hi ini Ml ii 1 ii in led with l4 C-, l5 N-, and 32 P- labelled ba< - 

terial cells. Soil Biol. Biochem. 19:559-565. 

Versteeg, M , and \ 1 udi I pi i, I"" '.. Partieipat in, testin f four low rental in nn tt . h 
n lo i i ddi nl i tilit; d I in vlono Pi inee. 1 nin j «i i i - ( 

Vine, H. 1953. Experiments on the maintenance of soil fertility at Ibadan, Nigeria, 1922-1951. 
Empire J. Exp. Agric. 21:65-85. 

von Uexkull, H.R nil E.\ kit i I Rehabilitation and lasting ii in i i i 1 1 I 

in Indonesia, p. 47-65. In Giessenei beitrage zur Entwiekliingsforsehung. Zentrum 
i nn i li ' i n Getmany. 

Weber, G, A. B.C. Robert, and R.J. Carsky. 1997. Handbook for use of LEXSYS (Legume Expert 
.t.nii Decision uppoi for integrating herbaeeou legum hit farming ystem UFA. 
Ibadan, Nigeria. 

W'etselaar. R. 1962. Nitrate distribution in tropical soils: 111. Downward movement and accumulation 
of nitrate in the subsoil. Plant Soil 16:19-31. 

W'etselaar. R.. and F. Ganry. 1982. Nitrogen balance in tropical agrosystems. p. 1-35. In Y.R. 
Dommergttes and H.G. Diem led. I Microbiology of tropical soils and plant productivity. 
Martinus Nijhoff/Dr W. Junk, the Hague, the Netherlands. 
in i I i I > i 1 h three tropical forage 

legumes in pure stands ami in grass mixtures. Agron. J. 59:47-59. 

Wild, A. 1972a. Mineralization ot soil nitrogen at a savanna site in Nigeria. Exp. Agric. 8:91-97. 

Wild, A. 1972b. Nitrate leaching under a bare fallow at a site in northern Nigeria. J. Soil Sci. 

Williams, TO., J.M. Powell, and S. Femandez-Rivera. 1995. Manure utilization, drought cycles and 
herd d\ mimics in die Sahel: Implications tor crop productivity, p. 393-199. In J.M. Powell et 
al. (ed.) Livestock and sustainable nutrient cycling in mixed farming systems of sub-Saharan 
Africa. Int. Livestock Ctr. lor Africa, Addis Ababa, Ethiopia. 

Wong, M.T.F., R. Hughes, and D.L. Rowell. 1990a. Retarded leaching of nitrate in acid soils from the 
tropics: me in n nt i i i i r, th nion lange capaci I >il i. 41:6 i 

Wong, M.T.F., R. Hughes, and D.L. Rowell. 1990b. The retention of nitrate in acid soils from the trop- 
ics. Soil Use Manage. 6:72-74. 


Wong, M.T.F., A. Wild, and A.S.R. Juo. 1987. Retarded leaching of nitrate measured in monolith 

lysimeters in south-east Nigeria. J. Soil Sci. 38:511-518. 
Wood, T.G. 1988. Termites and the soil environment. Biol. Fert. Soils 6:228-236. 
Woomer, P.L., A. Martin, A. Albrecht, D.V.S. Resck, and H.W. Scharpenseel. 1994. The importance 

and management of soil organic matter dynamics in the tropics, p. 47-80. In P.L. Woomer and 

M.J. Swift (ed.) The biological management of tropical soil fertility, lohn Wiley & Sow-.. 

Chichester, England. 
Wulijarni-Soetjipto, N., and R.F. Maligalig. 1997. Mucima pruriens (L.) DC. cv. group Utilis. p. 

199-203. In I. Faridah Hanum and L.J.G. van der Maessen (ed.) Plant resources of south-east 

Asia. Vol. I I. Auxilliary plants, Backhuys. Leiden, the Netherlands. 
Yamoah, C.F., A.A. Agboola, and G.F. Wilson. 1986. Nutrient contribution and maize performance in 

alley cropping systems. Agrofor. Syst. 4:247-254. 
Yoneyama, T, T. Murakami, N. Boonkerd, P. Wadisirisuk, S. Siripin, and K. Kouno. 1990. Natural 

l5 N abundance in shrub and tree legumes, Casuarina, and non N 2 fixing plants in Thailand. 

Plant Soil 128:287-292. 
Young, A. 1989 jrol ty foi oil i n > tion ( >iB I nl ^ illingford Englai I 
Zhang, H., M.L. Thompson, and J. Sandor. 1988. Compositional differences in organic matter among 

cultivated and uncultivated Argiudolls and Hapludalfs derived from loess. Soil Sci. Sue. Am. 

J. 52:216-222. 


Combined Use of Organic 
and Inorganic Nutrient Sources 
for Soil Fertility Maintenance 
and Replenishment 

Tropical Soil Biology and Fertility Programme 
Nairobi, Kenya 

Robert J.K. Myers 

International Crops Research Institute for the Semi-Arid Tropics 
Patancheru, Andhra Pradesh, India 

Stephen M. Nandwa 

Kenya Agricultural Research Institute 
Nairobi, Kenya 


The beneficial effects of combined organic and inorganic nutrients on soil fertility 
have been repeatedly shown, yet there are no guidelines for their management. Organic 
materials are not magic; many of their functions with respect to soil fertility are known. 
Organic materials influence nutrient availability (i) by nutrients added, (ii) through miner- 
alization-immobilization patterns, (iii) as an energy source for microbial activities, (iv) as 
precursors to soil organic matter (SOM), and (v) by reducing P sorption of the soil. The 
challenge is to combine orgames of differing quality with inorganic fertilizers to optimize 
nutrient availability to plants. Numerous field trials indicate both added benefits and dis- 
advantages of combining nutrient sources. Increased nutrient recovery and residual effects 
are associated with combined nutrient additions compared with inorganic fertilizers 
applied alone. Unfortunately, for many trials there is lack of crucial information on the 
nutrient content and quality of the organic inputs. Trials are needed that link the quality of 
the organic maten il to n i rtilizei :quivalency and its effect on the longer term composi- 
tion of SOM and crop yields. A systematic framework for investigating the combined use 
of organic and inorganic nutrient sources includes farm surveys, characterization of the 
quality of organic materials, assessment of the fertilizer equivalency value based on the 
quality of organics, and experimental designs for determining optimal combinations of 
nutrient sources. The desired outcome is tools that can be used by researchers, e 

Copyright © 1997 American Society of Agronomy and Soil Science Society of America, 677 S. 
Segoe Rd., Madison, Wl 5371 1, USA. Replenishing Soil Fertility in Africa. SSSA Special Publication 

194 PALM ET AL. 

ists, and farmers for assessing options of using scarce resource foi maintaining soil fertil- 
ity and improving crop yields. 

The impact of soil fertility replenishment projects i Sanchez et al., 1997, this pub- 
lication) is likely to be limited to a small number of farmers in the near future 
until appropriate programs and policies are put in place. More immediate strate- 
gies using farmer-available resources are needed that could reach more farmers, 
sooner. Although fertilizers are used in much of sub-Saharan Africa (hereafter 
referred to as Africa), the amounts applied are insufficient to meet crop demands 
(Smaling et al., 1997, this publication). Organic inputs are often proposed as 
alternatives to mineral fertilizers (hereafter referred to as inorganic fertilizers); 
however, the traditional organic inputs — crop residues and animal manures — 
cannot meet crop nutrient demand over large areas because of the limited quanti- 
ties available, the low nutrient content of the materials, and the high labor 
demands for processing and application. Most farmers in Africa fall within the 
two extremes of the organic to inorganic fertilizer continuum and use a combina- 
tion of organic and inorganic inputs (Table 8-1). Crop yields still fall short of 
their potential because of inadequate nutrient inputs, the inappropriate quality of 
the organic materials, and inefficient combinations. 

This chapter deals with the combined use of organic and inorganic nutrient 
sources for meeting crop demands and maintaining or restoring soil fertility. 
Given the high cost and uncertain accessibility of inorganic fertilizers in much of 
Africa, the goal should be to provide as much of the nutrients as possible through 
organic materials, making up the shortfall of the limiting nutrients through inor- 
ganic fertilizers. These goals would change as inorganic fertilizers became more 
affordable or available. The beneficial effects of the combined use of organic and 
inorganic nutrients on soil fertility, crop yields, and maintenance of SOM have 
been repeatedly shown in field trials, yet there are no predictive guidelines for 
their management, such as those that exist for inorganic fertilizers. 

The success of combined nutrient management depends on several factors, 
including the availability and affordability of different types of inorganic fertiliz- 
ers, the types and quantities of organic materials available, and the rates and pro- 
portions at which the two nutrient sources are combined. Most studies have 
included animal manures and crop residues as the organic additions while there 

Table 8-1. The use of nutrient inputs at selected locations in sub-Saharan Africa. 

Inorganic fertilizers 

Crop residues 

t Source, Bekunda at 


has been relatively little research on the use of alternative, higher quality organ- 
ic resources, such as leguminous cover crops and agroforestry species. 

A systematic approach that includes organic materials of different qualities 
and their combinations with inorganic nutrients is needed to develop guidelines 
for the selection and efficient management of these scarce resources. In the fol- 
lowing sections, the effect of organic materials on nutrient availability and crop 
production is reviewed from both scientific and practical perspectives. This 
review is followed by examples of combined nutrient management from field 
experimentation and modeling that could provide a framework for immediate 
field application and future research. 


Organic inputs can influence nutrient availability (i) by the total nutrients 
added, (ii) by controlling the net mineralization-immobilization patterns, (iii) as 
a source of C and energy to drive microbial activities, (iv) as precursors to SOM 
fractions, and (v) through interactions with the mineral soil in complexing toxic 
cations and reducing the P sorption capacity of the soil. In addition to these direct 
effects on nutrient availability, organic materials can affect root growth, pests, 
and soil physical properties that in turn influence nutrient acquisition and plant 
growth. The net effect of these different mechanisms on nutrient availability and 
plant growth differ with climatic regime, soil type, and quality and quantity of 
organic inputs. 

Sources of Nutrients 

Organic inputs such as manures, cover crops, and green manures have gen- 
erally been assessed in terms of their N concentration, while relatively little atten- 
tion has been paid to other macronutrients and micronutrients present. Organic 
inputs should be considered as complete fertilizers (N-P-K), perhaps the best 
being those containing or releasing the nutrients in the ratios and rates required 
by crops. 

Nutrient Contents 

The nutrient contents of organic materials, ranging from crop residues to 
agroindustrial wastes, vary widely. Table 8-2 compares the nutrients contained in 
a variety of organic materials with the nutrients required to produce a modest 2- 
t crop of maize (Zea mays L.) grain plus 3 t of stover. Although all the nutrients 
in the organic materials will not be available to a crop, the information can be 
used for an initial assessment of the type and amount of organic materials that are 
appropriate for a given cropping system and yield goal. These estimates can then 
be adjusted, knowing that crop recovery of N supplied by high-quality legumi- 
nous green manures is rarely more than 20% (Ciller & Cadisch, 1995) while that 
recovered from lower quality cereal stovers is generally much lower. 


Crop residues 

Maize stover 

Beau trash 

Banana leaves 

Sweet potato leaves 2 

Sugarcane trash 

Coffee husks 

Refuse coinpost+ 2 

Animal manures 

Cattle § 

High quality I 

Low quality 

Chicken ^ 

Farmyard chickenj ; 

Leguminous tree (leaves) 

Ctillitindra calotlixrsus : 

Gliricidia sepium : 

Leucaena leucocepha/a '. 

Sesbania sesban '. 

Senna spectabilis (non-N 2 -fixing) : 

Nonleguminous tree and shrubs (leaves) 

Chromolaena ordorata : 

Creviilea robusta 

Lantana camara \ 

Tithonia diversifolia : 

Leguminous cover crops 

Crotalaria ochroleuca * 

Dolichos lablab <■ 

kliH-ima priiriciis ': 

Nutrients required by 2 t maize grain + 3 t stover i 

I The TSBF database is the source of all data unless otherwise 

X Source, Nandwa (1996, unpublished data). 

§ Source, Mugwira and Mukurumbira, 1984. 

1 Source, Summers and Sutton. I9S0. not from Africa. 

Some organic materials such as poultry manures contain sufficient nutrients 
in 1 to 2 t for a 2-t maize crop, while others such as crop residues can require at 
least 10 t to match that contained in the 2-t maize crop. Cattle manure varies 
tremendously in its quality and fertilizer value. Extremes are found in manure 
obtained from commercial dairy farms compared with that from the communal 
areas of Zimbabwe (Mugwira & Mukurumbira, 1984). It is the latter, low-quali- 
ty manures, that predominate on smallholder farms of Africa (Probert et al., 
1995). In comparison, many leguminous trees and cover crops contain sufficient 
N in 2 or 3 t of leafy material (Giller et al., 1997, this publication). 

As a general rule, many organic materials when applied in modest amounts, 
i.e., <5 t dry matter ha" 1 , contain sufficient N to match that of a 2-t crop of maize 
but they cannot meet P requirements and must be supplemented by inorganic P in 
areas where P is deficient (Palm, 1995). 


Organic Materials, Adding or Recycling Nutrients? 

For organic materials to offset the nutrients removed from crop harvest, it 
is essential to differentiate between (i) organics produced on site and involved 
only in recycling of nutrients and (ii) organics produced elsewhere and carried to 
the site (biomass transfer), which count as an actual addition of nutrients. 
Recycling of nutrients on site may serve as a nutrient buffer to the system but will 
not redress the problem of nutrient depletion; nutrients are essentially being 
mined, except for cases of N2 fixation. 

On the other hand, biomass transfer is not widely practiced, except in cases 
where there is abundant animal manure available (Smith et al, 1993). Even the 
role of biomass transfer in redressing nutrient depletion must be evaluated at a 
scale larger than the plot because nutrients are being added to one location at the 
expense of another. For example, Swift et al. (1989) calculated that 96 kg N in 
manure, approximately lOt of low-quality manure, is needed to maintain 2 tha" 1 
maize yields. Expressed on an areal basis, 14 to 42 ha of miombo woodland graz- 
ing land are needed to supply that quantity of manure on a sustainable basis; at 
higher grazing levels the system is being mined and is no longer sustainable. The 
use of urban and agroindustrial wastes can reverse the flow of nutrients from crop 
harvest back to the farm, but this is probably not financially justified in most 
cases. A recent study in Kenya showed that 1 kg of N and P from compost costs 
US$ 0.50 and 1.20, respectively, compared with US$ 0.42 and 1.18 for N and P 
in purchased inorganic fertilizer (N = 200 g kg" 1 and P = 90 g kg" 1 ; S.M. Nandwa, 
1996, unpublished data). 

Quantities of Organic Materials Available to Farmers 

How do the amounts of organic materials needed to meet crop nutrient 
requirements compare with the amounts available to farmers? Average millet 
[Pennisetum glaucum (L.) R. Br.] stover production of 13 t ha" 1 in the Sahel is 
less than the 2 t of mulch recommended (Bationo et al., 1995). Crop residues also 
have competing uses, primarily as feed for livestock, that reduce the amounts 
available for managing soil fertility. An average of <700 kg ha" ' of manure, rang- 
ing from 450 to 1600 kg ha" 1 , is available for semiarid West Africa. This is much 
less than the 3 to 7 t ha" ' recommended for replenishing nutrients removed by 
crop harvest (Fernandez-Rivera et al, 1995). Manure production by zero-grazed 
cattle in Kenya has been estimated as 1 to 1.5 t animal" 1 yr- 1 (Strobel, 1987). Two 
animals would be needed to supply a 2-t maize crop, if the manure were of high 
quality, but eight animals would be required if the quality was low (Table 8-2). 

Increasingly the traditional nutrient sources for soil fertility management 
are produced in insufficient quantities and quality to meet crop demands. 
Alternative higher quality sources must be found, but there must also be niches 
on farms or the vicinity where they can be produced (Canity & Flinn, 1988). 
Leguminous plant materials provide higher quality organic inputs to meet N 
demands, if not P, but incorporating nonfood legumes in the farming system 
requires a sacrifice of space or time that is normally devoted to crop production. 
The additional labor requirement for planting, transporting, and incorporating 
these materials is also high (Ruhigwa et al., 1995). Therefore, farmers have not 

198 PALM ET AL. 

widely adopted planting legumes to improve soil fertility (Garrity & Flinn, 1988; 
Giller & Wilson, 1991). The economic and social trade-offs of improved soil fer- 
tility using legumes and other high-quality organic materials must be properly 
assessed in comparison with using crop residues and animal manures. 

Regulators of Mineralization-Immobilization Patterns 

Decomposition and nutrient release patterns are determined by climatic, 
edaphic, and resource quality factors (Swift et al., 1979). Of these factors, 
resource quality is most easily managed by farmers. Considerable research over 
the past century has related N release patterns to the resource quality, or chemi- 
cal characteristics, of organic materials (Heal et al., 1997). The N concentration 
and the C-to-N ratio of the material still probably serve as the most robust indices 
when all plant materials are considered (Constantinides & Fownes, 1994). 
Nitrogen concentration in tissue ranging from 18 to 22 g kg" 1 is the critical value 
for the transition from net immobilization to net mineralization. Not all organic 
materials with high N values, however, exhibit net N mineralization. Lignin con- 
tents >150 g kg" 1 slow N release considerably, and polyphenol contents >30 to 
40 g kg" 1 can result in net immobilization of N (Palm, 1995). Lignin and polyphe- 
nols are particularly important modifiers of N release for the fresh, nonsenescent 
leaves of high-quality materials (Constantinides & Fownes, 1994). The immobi- 
lization resulting from polyphenolics, particularly condensed tannins, may be 
much longer than the temporary immobilization resulting from high C-to-N ratios 
in cereal crop residues (Giller et al, 1997, this publication). 

Net P mineralization patterns are determined primarily by P concentration 
in the tissue. Materials with P content <2.5 g kg" 1 immobilize P (Blair & Boland, 
1978; Kwabiah, 1997). Phosphorus release patterns are not necessarily correlat- 
ed to N release. Some materials showing net N mineralization can result in net P 
immobilization and vice versa (Kwabiah, 1997, personal communication), stress- 
ing the importance of looking at more than N in organic materials. 

Traditional organic resources, primarily cereal crop residues and cattle 
manures, fall below the critical N content and immobilize N, at least temporari- 
ly. Tanner and Mugwira (1984) found that manures with N content <10 g kg" 1 
caused a decrease in the growth of maize seedlings for 4 wk that was related to 
immobilization. The negative effect of cereal residues on crop growth has been 
demonstrated in many field and pot trials (Ishuza, 1987; Nandwa, 1995). On an 
Alfisol in central Kenya, incorporation of maize stover reduced maize grain 
yields by 3 to 30% in the first three seasons. After the third year the reduction did 
not occur (Qureshi, 1987; Nandwa, 1995). Ishuza (1987) reported that incorpora- 
tion of 2.5 and 5.0 tha" 1 of stover resulted in 30 to 60% decreases in soil-avail- 
able N. 

Even if crop residues and other low-quality organic materials can be 
obtained in sufficient quantities, netN and probably P immobilization will occur, 
exacerbating the nutrient deficiencies, at least temporarily. The negative effects 
can be offset by combining with inorganic N (Msumali & Racz, 1978; Ganry et 
al, 1978; Paustian et al, 1992) or high-quality organic materials with N content 
>20 g kg" 1 and P >3 g kg" 1 (Smith et al., 1993). There are no guidelines as to the 


amounts of inorganics or high-quality materials needed to offset these negative 
effects, although as much as 100 kg N ha" 1 of fertilizer N was needed to over- 
come the immobilization resulting from mulching with maize stover in 
Guatemala (S. Waddington, 1997, personal communication). 

Although there is some degree of predictive capacity relating nutrient 
release patterns to organic resource quality, there has been little attempt to relate 
organic resource quality to fertilizer equivalency values. Frequent claims state 
that green manures have fertilizer equivalency values of 50 to 100 kg N ha" 1 
(Meelu & Morris, 1985; Ladha et al., 1988; Giller et al., 1997, this publication), 
but many trials do not provide sufficient information to allow relating the fertil- 
izer equivalency to the quality of the organic materials. Often there is no infor- 
mation given on the amount of the green manure added, its nutrient content, or its 
C constituents. Indeed as pointed out by Bouldin (1988), the green manure might 
be a replacement for a limiting nutrient, usually N, and in such cases a fertilizer 
equivalency value is useful. On the other hand, the green manure may have addi- 
tional nutrient or physical factors that influence the uptake and use efficiencies 
(Van Noordwijk & van de Geijn, 1996) and cannot be explained by the addition 
of N. In such cases, the fertilizer equivalency value is not very useful. 

Sources of Carbon and Energy for Soil Organisms 

Soil microbes can serve as sources and sinks of nutrients, and their activi- 
ty and turnover resulting from the decomposition of organic materials are con- 
sidered to be primary controlling factors in nutrient cycling and availability 
(Duxbury et al., 1989; Smith et al., 1993). Additions of organic residues can 
increase microbial pool sizes and activity, C and N mineralization rates, and 
enzyme activities (Smith et al, 1993), all factors that affect nutrient cycling. 
Since C is often the element most limiting to microbial growth and activity in, 
soils, the amount and metabolic activity, or C quality, of organic additions will 
influence rates of nutrient cycling.. Reinertsen et al. (1984) found that the size of 
the microbial biomass and rate of decomposition of wheat (Triticum aestivum L.) 
straw were determined by the size of the soluble C fraction of the organic mate- 

Additions of soluble forms of C also can result in the decomposition of 
more recalcitrant plant components and SOM, the so-called priming effect. 
Collins et al. (1990) found that the decomposition of mixes of wheat residues was 
greater than predicted when parts with more soluble C were added. Vanlauwe et 
al. fl994) also confirmed that more soluble C fractions in plant materials 
enhanced the decomposition of the more recalcitrant fractions. Other nutrients, 
particularly N and P, also are probably mineralized by the priming effect of solu- 
ble C, but the topic remains controversial (Azam et al, 1993). 

Organic additions to soil also can cause a shift in the distribution of nutri- 
ents in the organic or inorganic soil fractions caused by microbial activity. This 
redistribution might affect nutrient availability patterns and nutrient use efficien- 
cy, the net effect depending on the quality of the organic addition. As an exam- 
ple, Chauhan et al. (1981) and Hedley et al. (1982) measured changes in soil P 

200 PALM ET AL. 

fractions following addition of cellulose and N, plus or minus P. They concluded 
that for long-term build-up of soil P, it is necessary to add both a C and a P source. 
Carbon provides substrate for microbial growth, and subsequent microbial 
turnover results in the long-term accumulation of organic P, especially the more 
available P fractions. The same result may apply if a low-quality organic with low 
nutrient content but high amounts of available C was applied and resulted in net 
immobilization of the inorganic P. If higher proportions of N are held in soil 
organic fractions, they would be less susceptible to gaseous losses and leaching. 
On the other hand, additions of a high-quality material, with high nutrient content 
and high amounts of available C, may simply result in the substitution of the 
organic and inorganic sources of the nutrient (Jenkinson et al., 1985). 

In summary, C inputs, particularly the soluble fractions, modify the rate at 
which nutrients are cycled and become available and the form in which nutrients 
are held in the soil. This has not yet been translated into a clear framework for 
practical management, such as the amounts of different quality materials required 
to prime the system, or the proportions of organics (of different quality) to inor- 
ganics that would result in greater efficiency in the use of nutrients by reducing 
nutrient losses. 

Precursors to Soil Organic Matter 

It is through the formation of SOM that organic materials show longer term 
residual effects than do inorganic fertilizers. The use of inorganic fertilizers alone 
can even lead to a decline in SOM, while fertilizers combined with organics or 
organics used alone can maintain SOM levels (Sharpley, 1985; Bationo et al., 
1995). The relative roles of the quantity or quality of organic inputs in maintain- 
ing SOM, however, are not well understood. Many experiments include applica- 
tions of organic materials of different quality, but they also are applied in differ- 
ent quantities, making it difficult to interpret results. Even when similar amounts 
of different quality organic inputs are applied, some studies have shown that 
materials with higher C-to-N ratios and higher lignin contents result in more 
SOM (Janzen et al., 1988; Paustian et al., 1992), while others have shown no 
effect of organic input quality (Larson et al, 1972). 

Simply maintaining or increasing SOM may not necessarily lead to 
increased nutrient availability or productivity. Research in the past decade has 
focused on separating SOM into different fractions that are related to functional 
properties (Parton etal, 1989; Stevenson & Elliott, 1989), and particularly into a 
biologically meaningful fraction that is related to) nutrient-supplying capacity 
(Magid et al, 1996). Certain fractions, such as microbial biomass and the light 
fraction, have been positively correlated with N mineralization or N availability 
(Bonde et al., 1988; Hassink, 1995; Barrios et al., 1996). It is not yet clear how 
the quality of the organic input affects the different SOM fractions. Barrios et al. 
(1997) found that the amounts of light fraction under managed fallows of trees 
were higher for trees with low (lignin + polyphenol)-to-N ratios in litter than for 
trees with higher ratios in litter, although these results have been confounded with 
different amount of organic additions in the different treatments. 


Kapkiyai (1996) found in Kenya that additions of farmyard manure (FYM) 
over 18 yr increased the content and relative proportions of soil microbial bio- 
mass and particulate organic matter compared with additions of maize stover, 
though the amounts of FYM added also were much larger than those of maize 
stover. These more labile soil organic fractions were correlated to higher crop 

Organic amendments may increase SOM, depending on the amounts and 
quality of me materials added. Few experiments have controlled for the separate 
effects of the amount and the quality of the organic material and have included 
measurements of the resulting SOM fractions. Such experiments and measure- 
ments are necessary to identify possible relationships between organic inputs, 
SOM content and composition, and crop production. This information also is 
needed to determine how different types of organic materials produce residual 
effects in terms of a nutrient substitution value. 

Competitors for Phosphorus-Sorption Sites 

Organic materials have been shown to reduce the P-sorption capacity of the 
soil and increase P availability. The magnitude and duration of the effect varies 
with the soil type, the quality of the organic material, and the amounts added 
(Singh & Jones, 1976; Sivapalan & Sivasubramaniam, 1979; Bumaya & Naylor, 
1988; Iyamuremye et al., 1996a). In general, only materials with >2.5 g P kg" 1 
have been shown to reduce the P-sorption capacity (Singh & Jones, 1976). 

The mechanisms involved in this process are quite complex, as outlined in 
a recent review by Iyamuremye and Dick (1996). The most commonly cited 
mechanism refers to me action of organic acids produced from decomposition or 
root exudation. It is variously proposed that me organic anions (i) complex (or 
chelate) with ions of Fe and Al in me soil solution, preventing the precipitation 
of phosphate, and also reducing Al and Fe toxicity, (ii) compete with P for sorp- 
tion sites, and/or (iii) solubilize P from me insoluble Ca, Fe, and Al phosphates. 
Work by Hue (1991), Violante and Gianfreda (1993), and Staunton and Leprince 
(1996) indicate that complexation and competition are more important than 
replacing sorbed P or solubilizing native P. 

The most effective organic anions are the di- and tri-carboxylic acids such 
as tricarboxylic citric acid and me dicarboxylic malic, tartaric and oxalic acids, 
whereas monovalent acetate was found to have little effect (Hue, 1991; 
Iyamuremye & Dick, 1996; Staunton & Leprince, 1996). There is a question if 
organic acids can be found in the soil in sufficient quantities and for sufficient 
time to have such an effect on P availability. Levels of organic acids mat Hue 
(1991) and Staunton and Leprince (1996) added were similar to those found in 
soils. Although Staunton and Leprince (1996) report that these levels may not 
apply to me bulk soil, such levels may be found in the rhizosphere from root exu- 
dation. Iyamuremye et al. (1996b) found increased levels of malic, malonic, 
uccinic, formic, and acetic acid in soil following additions of manure, although 
the levels they added were quite high, >50 t ha" 1 . 

It is important to realize that most of the experiments showing reduction in 
P sorption with additions of organic materials have been conducted in the labo- 


202 PALM ET AL. 

ratory at unrealistically high loading rates. Another factor that has not be 
addressed adequately is the C quality of the organic materials and how it affects 
the production of different organic acids during decomposition. The practical 
issue then becomes, can organic inputs decrease P sorption and increase P avail- 
ability in the field under farmer circumstances at currently available types and 
rates of organic inputs? 

Indirect Effects on Nutrient Acquisition 

Organic materials also can have several other effects on soils and plants 
that influence nutrient acquisition and uptake by plants. Root growth can increase 
as a result of reduced exchangeable Al in the soil, caused by complexation with 
organic anions that are produced by me decomposition of organic materials 
(Hansen, 1989). It also can increase though an increase in pH caused by the addi- 
tion of basic cations from organic materials (Kretzschmar et al., 1991). Organic 
materials also can stimulate root growth either directly or through their effect on 
soil bacteria that can suppress root pathogens and produce plant growth hormones 
(Marschner, 1995). It is important to note that organics also can inhibit root 
growth, particularly ifphenolics concentrate in the soil or if bacteria detrimental 
to root growth increase because of the addition of organic materials. 

Applications of organic materials also can reduce or increase the numbers 
of pests and weeds, again depending on the quality of the material. Mulching with 
low-quality materials that decompose slowly has been shown to decrease weed 
biomass, while high-quality materials that decompose quickly have little effect 
(Fernandes et al., 1993; Salazar et al, 1993). The increased soil-water content 
resulting from mulch cover can, however, increase the incidence of pests. There 
is some evidence that the parasitic weed Striga sp. that reduces maize yields in 
much of Africa can be curtailed by applications of organic materials (Ransom, 
1996). The decrease is probably caused by several factors, including increased 
soil-water content, higher soil-available N levels, and perhaps even the suicidal 
germination of striga seeds caused by the products of organic decomposition 
(Vogtetal., 1991; Ransom, 1996). 

Soil physical properties such as structure, water content, and temperature 
can be affected by incorporation or surface application of organic materials. As 
an example, Tian et al. (1993) found that during drier periods lower quality 
mulches resulted in higher yields, as mineralization was probably higher because 
of the more favorable microclimate, lower soil temperatures, and higher soil- 
water content produced by the low-quality mulch. 

A summary of the role of organic materials in affecting nutrient availabili- 
ty and crop production brings out several points (Table 8-3). Despite some uncer- 
tainties in the role of organic materials, they are not magic. Nutrients from organ- 
ics, once mineralized, are no different from inorganic nutrients. Many of the func- 
tions of organics have been well described and understanding of them has 
reached a predictive stage. Now is the time to incorporate this predictive under- 
standing into the management of organic materials for soil fertility improvement. 


Factors affecting nutrient availability 
1. Nutrients added 

Factors affecting n 
6. Root growth 

Summary points 

• Insufficient quantity and quality of traditional organic 

sources to meet crop demand 

• Most materials recycle nutrients rather than add them to the 


• Immediate net N release it material has >18 to 22 g N kg" 1 
Lignin i i -nt 150 kg t :dw t n in i i iz iti ni itt 

luble polyphenol nteni U I i tilt in 

longed net N immobiltzatton 

• Net P release if material has >2.5 g P kg'" 

• C is limiting to microbial growth and activity in most soils 

• C addition can result in conversion of inorganic nutrients to 

organic forms 

» Metabolic C primes mineralization ofC N. ami P from 
more recalcitrant fractions 

■ Labile fraction, of SOM (microbial biomass and light frac- 
tion! arc correlated to nutrient availability 

• Relative effects of quantity and qualit) of organic materials 

on SOM and labile fractions not certain 

• Organic anions from decomposition or root exudation com- 

pete for P-sorption sites 

• P sorption reduced b\ materials with P >2.5 g kg'" 

• Effect of C quality of inputs not certain 

• Total effect on nutrient availability likely to be small 

• Al toxicity otsoi! reduced by organic materials, can result 

in increased root growth 
■ Organic materials stimulate (inhibit) root growth directh or 
mi ugh on in mi robi I popul tion 

• Low-quality mulches can suppress weed growth 

• Striga reduced with organic materials through effects on soil 

a r ,,i ]■ j 1 1 ni, iti i i of ti iga ed , 


The role of organics is varied and complex, as detailed above; the challenge 
is to use organics of differing quality in combination with inorganic fertilizers to 
optimize nutrient availability to plants. This requires knowing how the nutrient 
content and C quality of organic materials will add to and compensate for or will 
reduce nutrient availability from inorganic fertilizers. The term interaction is fre- 
quently used to describe the net effects of the combined use of organic and inor- 
ganic sources. This term implies to some a magic effect of organic materials, 
whereas to others it merely means a statistical interaction. A better phrase than 
interactions might be added benefits (or disadvantages) resulting from the com- 
bined use of organic and inorganic inputs compared with inorganics alone. In 
general, the nutrients supplied or removed (immobilized) by the addition of 
organics are additive to those supplied by inorganic nutrient sources (Paustian et 

204 PALM ET AL. 

al., 1992; Jones et al., 1996; Giller et al., 1997, this publication). Added benefits, 
or disadvantages, of combined nutrient additions are probably more related to the 
quality of the C substrate of the organic material and its effects on nutrient avail- 

Unfortunately, there has been little synthesis of the integrated effects of 
organic materials on net nutrient availability that can provide guidelines for com- 
bined nutrient management. An examination of past field trials and soil-crop 
simulation models, where the processes are integrated, may provide a starting 
point for assessing the relative importance of the various effects with different 
qualities of organic materials. 

The original intent of this chapter was to review and synthesize information 
from the numerous trials that have been conducted using combinations of organ- 
ic and inorganic nutrient sources. It quickly became apparent that most trials did 
not permit interpretation and extrapolation that would lead to management guide- 
lines because of their experimental designs. 

Numerous trials have compared the yields from a given amount of inor- 
ganic fertilizer (A), an organic material (B), and their combination (A + B), and in 
many situations (A + B) produced higher yields than A or B alone. It should not 
be surprising that the combination does better because more total nutrients have 
been added than in A or B alone. Nutrients from the organic sources are consid- 
ered in a additive manner (IA + IB) to that of the inorganic source, rather than 
looking at the substitution value (xA + yB, where the sum of x and y equals 1). 
Critical information on the nutrient content and quality of the organic material is 
often not provided. Despite some design flaws, a number of observations can be 
made regarding the combination of organic inputs of varying qualities and inor- 
ganic fertilizers on crop yields and nutrient-use efficiency. 

Synchrony of Nutrient Availability and Crop Demand 

A 4-yr experiment in India (Goyal et al., 1992) compared the N substitutive 
effects of wheat straw, FYM, and sesbania (species not given) green manure on 
pearl millet [Pennisetum glaucum (L.) R. Br.] yields, N uptake, and SOM. 
Nitrogen was applied to pearl millet at a rate of 120 kg N ha" 1 , as urea alone or 
one-half applied as urea and one-half applied as wheat straw, FYM, or sesbania. 
Data were presented for only the fourth year of the experiment (Table 8-4). Crop 
yields, N uptake, and N recovery were greater with the combination of FYM or 
green manure and urea compared with urea alone but less when wheat straw was 
combined with urea. The decrease in yields with wheat straw even after 4 yr is 
related to net N immobilization that would be expected from a material with a C- 
to-N ratio of 102. The authors attributed the higher N use in the combined sesba- 
nia or FYM with urea to the immediate availability of N from urea and its delayed 
release from the organics, achieving greater synchrony with crop demand. 

Another trial in India compared the N substitutive effect of organics by 
adding 80 kg N in various proportions of leucaena [Leucaena leucocephala 
(Lam.) de Wit] and urea, with 100, 50, 25, or 0% of the N added as leucaena and 
the remainder added as urea (Mittal, 1992). Applications were made annually for 
3 yr. Maize yields obtained from the 100% leucaena and 100% urea ti 


Table 8-4. Pearl millet yield, N uptake, N 


and soil 


following 4 

yr of application 

of fertilizer (urea) compared with the ( 



organic m; 

rterials of differing qualities 

(adapted from Goyal ■ 

atal.. 1992). 


Grain yield 

N uptake 




Microbial C 

tha- 1 

kg ha- 1 



mg C kg- 1 







N120 P40 






N60 P20 + N60 







N60 P20 + N60 

(wheat straw) 






N60 P20 + N60 







LSD (P = 0.05) 





t N = kg of N added as 

fertilizer or organic 

■ material. 


= kg 

of P added 

as inorganii 

; fertilizer. 

were similar the first 2 yr. Yields were slightly higher in the leucaena plots the 
third year. Yields from the 25% leucaena-75% urea combination were higher 
than the 100% urea or 100% leucaena treatments all 3 yr. 

Jones et al. (1997) compared maize yields and N-use efficiency in Malawi 
from applications of leucaena or gliricidia [Gliricidia sepium (Jacq.) Walp.] 
residues containing similar amounts of N. Residues were applied with or without 
inorganic N, but they were not compared with inorganic N applied alone. Yields 
andN use were higher for gliricidia residues than for leucaena residues. Gliricidia 
residues result in a large and rapid net N mineralization. Leucaena residues, with 
higher polyphenol content, exhibit initial net immobilization followed by net 
mineralization (Palm & Sanchez, 1991; Tian et al., 1992), the total N released by 
leucaena being less. Additions of inorganic N with the residues produced an 
increase in yields and N-use efficiency with leucaena but not with gliricidia. 
Jones et al. (1997) attribute the higher yields obtained from gliricidia to better 
synchrony of nutrient availability to crop demand. Addition of inorganic N to leu- 
caena improves synchrony by increasing the N supply at the initial stages of net 
immobilization resulting from applications of leucaena. Although this interpreta- 
tion is not necessarily incorrect, it is somewhat incomplete because yields from 
the sole application of equivalent amounts of inorganic N were not included for 
comparison, as was done by Goyal etal. (1992). 

Fertilizer N is subject to gaseous losses and leaching; losses of 20 to 40% 
are often reported (Van der Kruijs et al., 1988; Christianson et al., 1990). It is 
commonly believed that combining organics with inorganic fertilizer will 
increase synchrony and reduce losses by converting inorganic N into organic 
forms. Studies have generally looked at organic inputs of lower quality, such as 
crop residues. There are trade-offs between possible reductions in yields from the 
use of organic materials and greater potential nutrient losses with the use of inor- 
ganic nutrients alone. Is it possible that high-quality organic materials can reduce 
losses of inorganic N without reducing yields considerably? 

Janzen and Schaalje (1992) found that fertilizer N losses were twice as 
large when green manure plus fertilizer was applied to barley (Hordeum vulgare 


Table 8-5. Dry v, 
and K added as 


ts of 2-m 
green mi 


of 77//; 




in pots, 



ill equal amount of N, P, 

Nutrient added 





Weight of maize 

inorganic P 
r P 


Tithonia leaves 
N-P-K fertilizer 
Tithonia leaves + 
N-P-K fertilizer - 







4. 38 

LSD (P = 0.05) 

t Source, Gachengo, 1996. 

L.). Their interpretation was that green manure promoted high levels of nitrate 
and available C in the soil, enhancing denitrification. Losses were reduced with 
smaller repeated applications of green manure, again implying that the use of 
high-quality green manures as partial substitution for inorganic fertilizer N rather 
that addition to inorganic fertilizer may increase nutrient-use efficiency. Xu et al. 
(1993a,b) found large losses of 25 to 41% ofN added from leucaena prunings. 
They attribute this to denitrification. Losses are greater when materials are incor- 
porated rather than surface applied (Xu et al., 1993b; Jones et al., 1997). 
Although these studies did not compare the losses from fertilizer alone, they do 
indicate that losses from high-quality organics alone can be quite high. Ganry et 
al. (1978) also concluded that large applications of low-quality straw can result 
in large losses of fertilizer N through denitrification. These studies indicate that 
N losses can be quite large from both organic and inorganic sources, contrary to 
the popular belief that application of organic sources will result in fewer losses. 

Reduced Phosphorus Sorption 

Application of a high-quality organic material, tithonia [Tithonia diversifo- 
lia (Hemsley) A. Gray] leaves, combined with inorganic P in a pot trial resulted 
in greater maize biomass and P uptake than from equal amounts of nutrients 
added from inorganic fertilizers (Gachengo, 1996). The statistical analysis indi- 
cated a significant added benefit of the organic-inorganic treatment compared to 
the inorganic treatment (Table 8-5). A subsequent field study showed that titho- 
nia application reduced P sorption in the soil up to 16 wk (Nziguheba et al., 1998) 
and might account for the increased plant growth and uptake of P from the com- 
bined nutrient sources. 

Soil Organic Matter and Residual Effects 

In the study by Goyal et al. (1992) described above, the higher yield pro- 
duced by the sesbania plus urea and FYM plus urea compared with urea applica- 
tions alone also could reflect a residual effect after 4 yr of applying the organic 
materials. They found no differences in total soil C in plots receiving organic 


inputs compared with fertilizer alone following four annual additions of 2.7, 3.8, 
and 12 tha" 1 of sesbania, FYM, or wheat straw, respectively. Organic treatments 
did have higher levels of soil microbial biomass C and N, but yield differences 
from the different organic treatments (Table 8-4) were not related to the soil 
microbial biomass. The wheat straw plus fertilizer treatment had higher microbial 
C, relating more to the higher C input from wheat straw than to the quality of the 
inputs. This treatment also had the lowest yields. 

Increased Root Growth 

The effects of millet straw residue and/or inorganic fertilizer (N-P-K) on 
nutrient uptake, soil nutrients, exchangeable Al, and root growth were examined 
in Niger (Hafner et al, 1993). Higher P uptake rates in the millet residue plus fer- 
tilizer treatments was attributed to higher root-length density. This greater root 
growth could be a result of a reduction in exchangeable Al in the soil, or stimu- 
lation of root growth from the organic materials, or both. In an earlier study, the 
same research group found an increase in base saturation and pH and a decrease 
in labile Al with additions of millet straw (Kretzschmar et al, 1991) because of 
the addition of basic cations with the crop residue. The reduction of Al also might 
be partially due to chelation of Al by organic anions produced by the decompos- 
ing straw. The decrease in exchangeable Al was accompanied by an increase in 
extractable P. Large increases in N2 -fixing bacteria and total bacteria also were 
found in treatments with crop residues. As noted earlier, bacteria, particularly N2- 
fixing bacteria, have been shown to stimulate root production (Marschner, 1995). 

Results from these few trials indicate added benefits, and disadvantages, 
from the combination of organic and inorganic nutrient sources. While syn- 
chrony, nutrient-use efficiency, and residual effects of SOM are related to the 
addition of organic materials, N losses are not necessarily reduced. These added 
effects of combining nutrient sources are summarized in Table 8-6. The chal- 
lenge now is to attain a predictive capacity that integrates these processes and the 
influence of climate, soil, and organic input quality on these processes. 


Integrating the numerous and complex roles of organic materials in soil fer- 
tility and crop growth might be assisted by simulation models. Models, however, 
are only as good as the data that are used for constructing them, and knowledge 
gaps still exist on the role of organic materials of different qualities and their 
interactions with inorganic fertilizers in modifying soil nutrient availability. More 
controlled experiments are needed that can be used for developing and validating 
models. Once these models exist, they can assist in evaluating organic-inorganic 
combinations and in selecting and optimizing combinations. 

Models that can be used for investigating the combined effects of organic 
and inorganic inputs must have the capacity for (i) additions of organic inputs of 
differing quality, with decomposition routines that include organic quality para- 
meters, (ii) links between organic quality and the formation of different SOM 


Potential benefit or disadvantage ol combined nutrient sources 

I. Organic materials provide addition, il nutrients that may be colimiting 
to plant growth 

2. Organic materials alter nutrient availability patterns 

A. High-quality materials (low C/N ratio, low lignin. low polyphenol) 
produce somewhat delayed N availabilit) patterns, when coupled 
with ini rg mi fertilize! in i i I lemand i i i 

B. Medium-quality materials (low C/X ratio, low lignin. high polyphenol), 
may decrease N availability over the short- and medium-term if 
polyphenols have high protein-binding capacity but may result in 
greater residual effect in the longer term 

C. Lo\ i matt I i I « ratio, high ul I il i 
immobilization of added fertilizers 

[). Low-quality materials (high lignin) may decompose faster when 
combined with inorganic N 

3. Available C from organic* primes the mineralization of N and P from soil 
organic matter, adding to the available nutrients 

4. Organic materials reduce P sorption b\ mineral soil, making added inorgani 
P more available 

5. Organic materials build up soil organic X and P in the longer term and 

h ii lie i h ii i n 1 Ih i it t i in hi li i i 

6. Organics increase root growth so plants are better able to exploit fertilizer 

7. Organics reduce weed populations and reduce competition for added fertilize 

= neutral or small impact; the number of + 

fractions, (iii) crop production models that are sensitive to short-term changes in 
nutrient availability, and (iv) both N and P dynamics, which in the case of P must 
include P sorption as affected by organic materials. The strengths and weakness 
of the various models that currently exist are reviewed below. 

Models vary considerably in their ability to handle different quality organ- 
ic inputs. The CERES models include decomposition and N dynamics of cereal 
and legume residues, with decomposition determined by the C-to-N ratio of the 
organic material (Dimes, 1996). In the CENTURY model, decomposition and N 
mineralization are driven primarily by the lignin-to-N ratio of the material 
(Partem et al, 1989). Recently, Whitmore and Handayanto (1997) included the 
effect of polyphenolics and their protein-binding capacity on N dynamics and 
availability, which is particularly important for systems that include legumes. 

Models of P in soil-plant systems initially were physicochemically based, 
with organics and P cycling ignored. Currently, models such as EPIC (Jones et 
al., 1991) and CENTURY include organic P components. The CENTURY sub- 
model divides P into several pools, both organic and inorganic, with a primary P 
pool (assumed to be apatite), secondary, occluded and labile mineral P pools. 
Recent users of the CENTURY model have been rather critical of the P submod- 
el for tropical soils (Gijsman et al., 1996). None of the models adequately deal 
with P sorption and the effects of organic additions. 


The strengths of the CENTURY model are its ability to predict long-term 
changes in SOM; however, it does not simulate short-term nutrient availability 
and crop production well because of the month-long time step (Paustian et al., 
1992). Cropping systems models, such as APSIM (McCown et al, 1996) and 
DSSAT (Tusji et al., 1994), simulate short-term crop production but are limited 
in their ability to model decomposition and SOM dynamics. Thus far, there seem 
to have been no attempts to merge the detailed soil process models with the crop 
models that are better at simulating crop growth and yield. Van Noordwijk and 
Van de Gejin (1996) also stress that models mat include root processes are essen- 
tial for simulating nutrient dynamics. 

Myers (1995) used CENTURY to simulate a 15-yr experiment conducted 
in Thailand in which a range of organic materials was applied in the presence and 
absence of N fertilizer. The model successfully provided a qualitative description 
of the changes in SOM but underestimated SOM accumulation in soil, particu- 
larly in plots that received fertilizer and crop residues. Paustian et al. (1992) 
reported similar discrepancies. The model also underestimated SOM in plots that 
received the largest amounts of organic inputs. Aboveground crop production 
was simulated well in only about one-half the cases and corresponded to cases 
where SOM was simulated well. 

In the opinion of these authors, current simulation models do not yet fully 
meet the needs of research and extension workers in developing countries. The 
major issues that need attention, assuming that the suite of crop models is ade- 
quate, are the capacity to simulate P dynamics and the decomposition of a broad- 
er range of crop residues and higher quality organics that are likely to be encoun- 
tered in tropical farming systems. Development of models that can be used in 
such situations for selecting appropriate mixes of organic and inorganic inputs 
will require closer interaction between the modelers and the experimentalists. 


The capacity to make practical recommendations on the use of organic 
materials as a source of nutrients for crops will be limited until the different fac- 
tors that affect yields are separated and accounted for through appropriate trials 
and the organic resources are sufficiently characterized. This is a complex task. 
To advance from the current empirical status to a predictive capacity for the selec- 
tion and management of combined organic and inorganic inputs, a few knowl- 
edge gaps must be filled, which requires appropriate hypotheses and experimen- 
tal designs. A systematic framework is proposed for the investigation of com- 
bined organic and inorganic inputs (Fig. 8-1). 

An initial step is to conduct surveys that include interviews and visual 
assessment of the farm vicinity and community, on the availability and quantities 
of organic materials and fertilizer. Emphasis should be placed on alternative, 
high-quality materials that exist in the landscape or for niches to plant nonfood 
crop legumes. The surveys should also assess the farmers' current soil fertility 
management practices, constraints, and opportunities. The soils of the area also 


Availability and types of inorganic fertilizers 
Availability and amounts of organic inputs 

Laboratory incubs 

Relationship to organic input quality 

Factorial trials 
Long-term and residual e' 

must be characterized in terms of the primary limiting nutrients. This can be done 
by soil tests or limiting nutrient trials. 

Organic materials that have been identified should then be characterized for 
their quality parameters. The minimum of parameters that should be included are 
macronutrient concentrations, lignin, soluble C, ash, and soluble polyphenols (if 
the N concentration is >18 g kg" 1 ). Standardized methods for these analyses are 
also recommended (Palm & Rowland, 1997). Once a sufficient number of mate- 
rials of the same species have been characterized to indicate the possible range in 
quality within a species, it might be possible to categorize a plant material into a 
specific quality grouping without analyzing the material. 

Fertilizer equivalency or nutrient substitution values of organic materials 
can then be determined and related to the quality of the material. Such informa- 
tion could be obtained through a combination of laboratory incubations and field 
trials. Incubations establish the amount of different organic materials needed to 
attain similar soil available nutrient levels for a given amount of fertilizer. Field 
trials test recommendations from the incubations on different soils and climates, 
and models extrapolate to other types of organic materials and environments. 

Field trials usually relate the yield obtained from organic inputs to the 
yields obtained from an inorganic response curve. One must be certain of the lim- 
iting or colimiting nutrients of a particular soil and then decide if the trial will 
assess the nutrient equivalency of one or multiple nutrients. If only one n 


is to be assessed then the other nutrients must be supplied in nonlimiting quanti- 
ties in the inorganic response curve and the organic treatments. If multiple nutri- 
ents are to be assessed, additional multiple nutrient response curves must be 
included. Rates of nutrients applied from the organic sources must be on the 
responsive part of the fertilizer response curves; if not, then the trials will not be 

The outcome of these types of trials will permit grouping of organic mate- 
rials into like categories, based on their quality, that have similar nutrient substi- 
tution capacities per unit of added material. As an example, the recommendation 
might say that for materials with N content of 40 g kg" 1 , lignin content <50 g 
kg" 1 , and polyphenol <30 g kg" 1 , then 4 t of material is needed to produce an N 
equivalency of 80 kg N. This type of material containing 160 kg N would have 
an N-application efficiency 50% that of the fertilizer. These recommendations 
would, of course, differ with soil, climate, and crop. 

Once fertilizer equivalency values have been established for different 
groups of plant materials, trials can determine the substitutive effect of different 
quality organics at different proportions of organic -to-inorganic sources. These 
trials could be factorial with several rates of both the organic and the inorganic 
materials or substitutive in which the organic and inorganic inputs are added at 
different proportions but the total amount of nutrient added is the same (Mittal et 
al, 1992). Both types of trials have merit. 

A factorial arrangement of treatments provides a means for comparing dif- 
ferent rates but has the limitation that most of the combined treatments are addi- 
tive rather than substitutive in nature. The number of treatments quickly becomes 
prohibitive if more than one organic material is assessed, in which case con- 
founded designs can be used (Jones, 1996). For assessing low-quality materials, 
perhaps only one rate that is normally used by farmers should be used with sev- 
eral rates of the inorganic to determine how much inorganic fertilizer is needed 
to overcome the negative effect of the low-quality organic. 

The other design, in which the total amount of nutrients added is the same 
but the proportion of organic -to-inorganic source changes, is useful to determine 
the optimal combinations in terms of economics, nutrient-use efficiency, and 
residual effects. This design will indicate if the two nutrient sources are merely 
substitutes (Ql) or if there is some additional benefit (Q3) or disadvantage (Q2) 
to be derived from the organic material (Fig. 8-2). Economic analyses of the var- 
ious organic-inorganic combinations should be conducted under current condi- 
tions and future scenarios, to indicate realistic management options. 

These trials should be planned for the long term to assess residual effects 
and changes in SOM composition as they relate to the quality of the organic mate- 
rial and the proportion of organic to inorganic. Until such trials are conducted that 
link the quality of the organic material to its fertilizer equivalency value and its 
effect on the longer term composition of SOM and crop yields, there will be no 
means of providing guidelines for the combination and efficient use of organic 
and inorganic inputs. 

The desired outcome of the research process detailed above is to acquire 
fairly simple tools that can be used by researchers, extensionists, and farmers for 
assessing different ways of using scarce resources for maintaining soil fertility 



Quantity of organic material added 

Fig. S-2. A conceptual mode! depicting the amounts and proportions of organic materials of differing 
qualm iQI ()' () i i I inoi n , tili n ry to achi i red yield -\ll lines repre- 

sent the same yield. FEQ1, FEQ2, and FE« i pi nt tl imom i I i e lespective organ- 

l I 1 I I 1 111! || || I |1 Ml tili i\ I I 

and improving crop yields. An example of one such tool is a preliminary decision 
tree on the uses of organic material of different qualities forN management (Fig. 
8-3). The decision tree is a current best bet based on research results. It can be 
modified as more information becomes available, but more importantly it can be 
implemented with farmers and modified based on their experiences and available 

1 Organic materials available 



Many farmers in Africa, and their colleagues worldwide, are currently 
combining organic and inorganic sources of nutrients to try to meet crop 
demands. Yields obtained are far below their potential because of inadequate 
amounts added, the low quality of the organic materials, and inappropriate or 
inefficient combinations. Given the high cost and uncertain accessibility of inor- 
ganic fertilizers in Africa, the goal should be to provide much of the nutrients 
through organic inputs, making up the shortfall of the limiting nutrients through 
inorganic fertilizers. Where the quantity and quality of organic materials is low, 
there is a need to find alternative, high-quality organic materials that can be incor- 
porated in the current farming systems. 

Despite numerous field trials combining organic and inorganic inputs, it is 
not at present possible to recommend guidelines for combining organic and inor- 
ganic nutrients because of inadequate experimental design and little information 
on the quality of organic inputs. Inorganic fertilizers can offset the negative 
effects of low-quality organics but how much cannot be specified. Higher quali- 
ty organic materials can substitute for inorganic fertilizers, but there are no pre- 
scriptive guidelines that relate the quality of the organic material to its nutrient 
substitution value. 

Prescriptive guidelines can be attained once links are made between the 
quality of the organic materials and their short-term fertilizer equivalency value 
and longer term residual effects through SOM formation. These guidelines also 
must incorporate farmer perceptions and circumstances, including available 
resources, resource allocation, farm niches, soil types, and limiting n 


The authors would like to thank Roland Buresh, Rob Gilbert, Roel Merckx, 
Meine van Noordwijk, and Stephen Waddington for their valuable comments. 
Ken Giller and Sieglinde Snapp are especially thanked for their thoughts and dis- 
cussion about the interactions of organic and organic nutrient sources. 


Azam, F., F.W. Simmons 

Barrios, E., R.J. Buresh, and J.I. Sprent. 1996. Nitn i mineral] < n n d u it) fractions of soil 

organ matter from mai md legu ppii ystem n ' ' i i ! ' I l 3. 

Barrios. E.. F. Kwesiga. R.I. Buresh. and J.I. Sprent. 1997 Light fraction soil organic matter and 
I i I n ii n following ti m i i ii i \ n ' ' I I 

Bationo, A., A. Buerkert, MP. Sedogo, B.C. Christianson, and A.U. Mokwunye. 1995. A critical 
review of crop-re ichi u i il inn Inn n n th " i Vf'rii n mi-arid tropii p. 
30.3-322. In . I. M. Powell et al led.l Livestock anil sustainable null ient cycling in mixed farm- 
ing - ' in or in- ilia n Urica Vol. II. Tech. Pap. Int. Livestock Ctr. for Africa, Addis 
Ababa, Ethiopia. 

Bekunda, M.A., and P.L. Woomer. 1996. Organic resource management in banana-based cropping 
systems ofthe Lake Victoria basin. Uganda. Agric. I t. Envh 171- ISO 


e.rial added to soil. Aust. J. Soil 

Bonde, T.A., J. Schnurer, and T. Roswall. 1988 Miu bi il bi n i i fra ti n ot potentially min- 
eralizable nitrogen in soils from long-term field experiments. Soil Biol. Biochem. 20:447-452. 

Bonklin. D.R. I98S. Effect of green manure on soil organic matter content and nitrogen availability, 
p. 151-163. /;/ Green manure in rice farming. The role of green manure crops in rice tanning 
systems. Int. Rice Res. Inst, Los Banos, Philippines. 

1 i 1 1 i II ncl D It i I 'hiii r] ii ii i tractabilitv in And il incubal 

ed with plant residues of variable P content. Plant Soil 1 12:77-81. 

Chauhan, B.S., J.W.B. Stewart, and E.A. Paul. 1981. Effect of labile inorganic phosphate status and 
oriianic carbon additions on the microbial uptake ot phosphorus in soils. Can. ,1. For. Res. 

Christianson, C.B., A. Bationo, J. Henao, and P.L.G. Vlek. 1990. Fate and efficiency of N fertilizers 
applied to pearl millet in Niger. Plant Soil 125:221-231. 

Collins, H.P., F.L. Elliott, R.W. Rickman, D.F. Bezdicek, and R.I. Papendick. 1990. Decomposition 
no mi i ii M i ii; ir id i , m ii in in ul ' i i Mn I s4 780-785. 

Constaninides, M.. and J.H. Fownes. 1994. Nitrogen mineralization from leaves and litter of tropical 
1 i Relationship to niti i. lignin ai III Ivphenol ntratioi il Biol 

Biochem. 26:49-55. 

Dimes. J. 1996. Simulation of mineral N supply to no-till crops in the semi and tropics. I'll D. thesis. 
Griffith Univ.. Queensland Australia. 

Duxbury, J.M., M.S. Smith, and J.W. Doran. 1989. Soil organic matter as a source and a sink of plant 
nutrients, p. 33-68. /// D.C. Coleman et al. (ed.l Dynamics of soil organic matter in tropical 
ecosystems. Univ. ot Hawaii Press, Honolulu. 

Fernandes, E.C.M., C.B. Davey, and L.A. Nelson . 1993. Alle\ ci | iin in m u id soil in the Upper 
Mn ' ui mull Ii rtiliz i ind h I i ro\ n ot pit nil ; fects. p 7 96 In I R; gl n in 1 1 
( (ed.l Technol foi tainabl riculture in the troj Publ MSA 

CSSA, and SSSA, Madison, WI. 

Fernandez-Rivera, S., T.O. Williams, P. Hiernaux, and J.M. Powell. 1995. Fecal excretion by rumi- 
nants and mam i r iila ilit; foi rop pi duction in emi-arid \V t i U'ri i , 149-lt u 
111 II I I L I II 1 1 It 111 1 1 1 II 1 til 

of sub-Saharan Africa. Vol. II. Tech. Pap. Int. Livestock Ctr. for Africa, Addis Ababa. 

Gachengo, C N I99i Pin ph u i Ii ind e ul ilit; n klition 1 »i mi materials to phos- 
phorus fixing soils. M.Phil. thesis. Moi Univ.. Eldoret, Kenya. 

Ganry. F.. (.1. Guiraud. and Y. Dommergues. 1978. Effei t or straw incorporation on the yield and nitro- 
gen I I mi in ill ind\ iil-pt 1 1 i lill t cropping system o1 negal. PI m oil ' 662 

Canity, D.P., and J.C. Flinn. 1988 Farm-le i 1 m m in. nt tin for g e n in mine ciops in Asian 
rice environments, p. 111-129. In Gi en manui n ri i rmin Ih rol ol i n n mi 
crops in rice farming systems. Int. Rice Res. Inst.. Los Bancs. Philippines. 

Gijsman, A.J., A. Oberson, H. Tiessen, and D.K. Friesen. 1996. Limited applicability of the CEN- 
TURY model to highly weathered tropical soils. Agron. J. 88:894-903. 

Ciller. I id G. Cadisch. I Future benefits from biological ni n fixation: An ecological 

approach to agriculture. Plant Soil 174:255-277. 

Giller, K.E., G. Cadisch, C. Ehaliotis, E. Adams, W.D. Sakala, and P.L. Mafongoya. 1997. Building 
soil nitrogen capital in Africa, p. 151-192. In R.J. Buresh et al. ted.l Replenishing soil fertil- 
ity in Africa. SSSA Spec. Publ. 51. SSSA, Madison, WI (this publication). 

Giller, K.E., and K.J. Wilson. 1991. Nitrogen fixation in tropical cropping systems. CAB Int. 
Wallingford, England. 

Goyal, S., M.M. Mishra, I.S. Hooda, and R. Singh. 1992. Organic matter-microbial biomass relation- 
ships ui i, id X] rim i ni I i tropii I ondition Effect i1 inoi mi i tili tion and 
organic amendments. Soil Biol. Biochem. 24:1081-1084. 

Hafner, H., E. George, A. Bationo, and H. Marschner. 1993 Effi ol oj i. nines on root growth 
and phosphorus acquisition ot near! millet in an acid sandy soil in Niger. Plant Soil 

Hansen. .1. W. 1989. influence of decomposing organic material on aluminum chemistry and phyto- 
toxicity. Ph.D. diss. Univ. of Hawaii, Honolulu. 

Hassink. J. 1995. Decomposition rate constants ot size and density fractions of soil organic matter. 
Soil Sci. Soc. Am. J. 59:1631-1635. 

Heal. O.W.. J M. Anderson, and M.J. Swift. 1997. Plant litter quality and decomposition: An histori- 
cal overview, p. 3-30. In G. Cadisch and K. Culler (ed.l Driven by nature: Plant litter quality 
and decomposition. CAB Int., Wallingford. England. 


Hedley, M.J., J.W.B. Stewart, 

tions induced hv culth 

i!i 1991 . El I uganii ids/anions on P si rptioi in i ph I i nl il ilil in n ii i lil 

ferent mineralogies. Soil Sci. 152:463-471. 
Ishuza, I.L.K. 1987. The effects of applying maize straw to sc 

izer N and P and on maize yield. M.Sc. thesis. Sokoins 
lyamuremye, F.. and R.P. Dick. 1996. Organic amendments and phosphorus sorption by soils. Adv. 

Agron. 56:139-185. 
lyamuremye, F.. R.P. Dick, and J. Baliam. 1996a. Organic amendments and phosphorus dynamics: I. 

Phosphorus chemistry and sorption. Soil Sci. 161:426-435. 
lyamuremye, F., R.P. Dick., and J. Baham. 1996b. Organic amendments and phosphorus dynamics: 

III. Phosphorus speciation. Soil Sci. 161:444-451. 
Jan n. H.H ind G.I haul j I ' T ul pon i nil n nd non-i utriti nal benefit i 

legume green manure. Plant Soil 142:19-30. 
Janzen, R.A., C.F. Shayekwich, and Goh Tee Boon. 1988. Stabilization of residual C and N in soil. 

Can. J. Soil Sci. 68:733-745. 
Jenkinson, D.S., R.H. Fox, and J.H. Rayner. 1985. Interactions between fertilizer nitrogen and soil 

nitrogen — the so-called 'priming' effect. J. Soil Sci. 36:425-444. 
I C'.A h j ill l ii 1 '1 I i i h u I i i i 1 1 I j I in 

system p 32 /;/ .1. Hani ind J.T. Ritcl i loci lin [I in n soil systems. ASA 

Agron. Monogr. 31. ASA, CSSA, and SSSA, Madison, WI. 
Ion II Efficienl mental ns foi r ir i I it rch i k meth ^ ' in 

il I i In 1 i 1 F i n i ii Mi 

Africa. CIMMYT, Harare, Zimbabwe. 
Jones. R.B.. S.S. Snapp. and H.S.K. Phombeya. 1997 Management of leguminous leaf residues to 
I /(111! 

Cillei led. i Driven by nature: Plant litter qiialit) and decomposition. CAB Int.. Wallingford. 

Jones, R.B., J.W Wendt, \\ T I in rson id O ' Inn i 1996 I u n i - • ii lize alley cropping in 

Malawi: 1, Effects of N. P. md lea ppli ill i n mai vi Id ind i il iropt tii \iirofoi 

Syst. 33:281-294. 
pkiyai I 1996. Dynamics of soil organic carbon, nitrogen, and microbial biomass in a long-term 

experiment as affected by inorganic and organic fertilizers. M.Sc. thesis. Univ. of Nairobi, 

Nairobi, Kenya. 
Kretzschmar, R.M., H. Hafner, A. Bationo, and H. Marschner. 1991. Long- and short-term effects of 

• i '[i 1 1 on aluminum toxical pliosphon availabilit nd growth of \ u I millet in an 

acid sandy soil. Plant Soil 136:215-223. 
Ladha, J.K., I. Watanabe, and S. Saono. 1988. Nitrogen fixation b\ leguminous green manure and 

practices tor its enhancement in tropical lowland rice. p. 165-183. In Green manure in rice 

farming. The role id green manure crops in rice tanning systems. Int. Rice Res. Inst., Los 

Bancs. Philippines. 
Larson, W.E., C.E. Clapp, W.H. Pierre, and Y.B. Morachan. 1972. Effects of increasing amounts of 

organic i in n cm til uou nil 11 < n • mil i irbon niti >i n pin ph m nd ulj hi i 

Agron. J. 64:204-208. 
1 gid I \ oi n nd I I i ill , I 1 i( lii i.iiiii Ii i ii i action of soil organ- 

ic matter: Three size-density fraction methods tor tracing the fate of homogeneously 4 C- 

labelled plant materials. Soil Biol. Biochem. 28:89-99. 
Marschner, H. 1995. Mineral nutrition of highei plants. 2nd ed. Academic Press, San Diego, CA. 
McCown, R.L., G.L. Hammer, J.N.G. Hargreaves, D.P. Holzworth, and D.M. Freebairn. 1996. 

APSIM: A novel software system for model development, model testing and simulation in 
ii ulturai i in i > me Syst. 50:255-271. 

Meelu. O.P.. and R.A. Morris. 1985. Integrated management of green manure, farmyard manure, and 
i iii a id ii i t I i i i i i ii i p 185-193. In 

Efficiency of nitrogen fertilizer tor rice. Int. Rice Res. Inst., Los Bains. Philippines. 
Mittal, S.P., S.S. Grewal, Y. Agnihotri, and A.D. Sud. 1992. Substitution of nitrogen requirement of 

maize through leaf biomass of Leucaeiui li'iicncephahi: Astronomic and economic considera- 
tions. Agrofor. Syst. 19:207-216. 
Msumali, G.P., and G.J. Racz. 1978. The use ofN ls in nitrogen immobilization studies in the soil. 
i I I i 1 i i 1 t h iiii I i i i ! in 

ofthe added nitrogen. E. Afr. Agric. For. J. 43:232-237. 

216 PALM ET AL. 

Mugwira, L.M., and L.M. Mukurumbira. 1984. Compai ' I ti I ui mines from the com- 

munal areas and commercial feedlots as plain nutrient sources. Zimbabwe Agric. J. 

1 til I I I Fl I ) M ! I III III III 

p. 1 31- 14 / t Powell et al. led.} Li to ml tainabl nutriei lin in mi <J 

i in i in il li i in III 1 [ j li L t il i i I 

Ababa, Ethiopia. 
Myers, R.J.K. 1995. Modelling of soil organic matter dynamics, p. 140-148. In R.D.B. Lefroy et al. 

(eel. I Proi .t l i ' li | ii ! * » ni Iattei Iana ment foi ustaii il ricultui 

Ubon, Thailand. 24-26 Aug. 1994. ACIAR Proc. 56. ACIAR, Canberra, Australia. 
i I 11 i i i i i nil t M ke through in nze t ver 

placement and N teitil iti n mil ntinuou niaiz mono n iping ^ stems in Kenya. 

Ph.D. thesis. Univ. of Exeter, Exeter, England. 
Nziguheba, C, C.A. Palm, R.J. Buresh, and P.C. Smithson. 1998. Soil phosphorus fractions and 

adsorption as affected b\ organic and inorganic sources. Plant Soil (in press). 
"aim. C.A. 1 < i ii n ii i ii nutrient i irement I i i i i [ , [ I nt 

Agrofor. Syst .30:105-124. 
,'i in C.A in I [' Ro land 199 " h mi il naci ri u i , pi ni , dip toi decomposition. 

p. 379-392. In G. Cadisch and K.E. Giller (ed.) Driven by nature: Plant litter quality and 

decomposition. CAB Int., Wallingford. England. 
Palm, C.A., and P.A. Sanchez. 1991. Nitrogen release from the leaves of some tropical legumes as 

i i u i i 'iii in i I i I [ uli in 1 1 1 1.1 ii i i 

Parton, W.J., C.V. Cole, J.W.B. Stewart, D.S. Ojima, and D.S. Schimel. 1989. Simulating regional pat- 
terns of soil, C. X. and P dynamics in the U.S. central grasslands region, p. 99-108. In M. 

< I uli lin ill II Eiei ii >m i d i I ilo in ii i ili I ind P » li intl h IK ng Kluw i 

Academic Publ.. Dordrecht, the Netherlands. 
Paustian. K.. W..I. Parton. and .1. Persson. 1992. Modeling soil organic matter in organic amended and 

nitrogen-fertilized long-term plots. Soil Sci. Soc. Am. J. 56:476-488. 
Probert, M.E., B.A. Keating, J.P. Thompson, and W.J. Parton. 1995. Modelling water, nitrogen and 

crop yield for a long-term tallow management experiment. Aust. J. Exp. Agric. 35:941-950. 

■ ' i ' in I 19! I I, cumulatp r i * i u -P fertilize] manui mid op lesidues on maize 

rain yield leaf nutrient conten n I i il i i I il ete In National 

1 i' agronomy Workshop, 17-19 Feb. 1987. Kenya Agric. Res. Inst, Nairobi, Kenya. 

1 1 i I i lin i I i i in i i null i n llu f sul i in i^ii | 

623-288. In I he 2nd lin Control C'ongi Copenh n, Denmarl lun 

I i I ' i i I * - i • i I m i [ nil L i I I i I ijerg I I i 1 i mi, il 

Reinertsen, S.A., Elliott L.F., Cochran V.L., and Campbell G.S. 1984. Role of available carbon and 

nitrogen in determining the rate 01 straw decomposition. Soil Biol. Biochem. 16:459-464. 
RuhigwaBA. \I P i liuru. [ < ncer. and F nn n I ' nomi analysis of cut-and- 

ii nd li' i ppin s iii i i m i n "In in i for pi in nn in oi theastern Ii 1 1 

UTARes. 11:11-14. 
i i ii , L I i ii \ ' li i I i i | ii i i ii ii m id ding s> tems on allu- 

vial soils of the Upper Amazon Basin. Agrofor. Syst. 22:67-82. 
Sanchez, PA., K.D. Shepherd, M.J. Soule, F.M. Place, R.J. Buresh, A.-M.N. Izac, A.U. Mokwunye, 
F.R. Kwesiga, C.G. Ndiritu, and PL. Woomei 1 il ferti plenishment in Africa: An 

II II I HI 1 1 lU I 1 I 1 I 1 | I -T I I [ill 11 I ll I I I Hi 111, I 1 | 

tility in Africa. SSSA Spec. Publ. 51. SSSA, Madison, WI (this publication). 

i I II yh i I n i i i i iih I i i oh ii in i i soils Soil Sci. Soc. 

Am. J. 49:905-911. 
Singh, B.B., and J.P. Jones. 1976. Soil fertility and plant nutrition. Soil Sci. Soc. Am. J. 40:389-394. 

Si\ ij d in k nd i ii I I in I il | l I 1 i I 1 I 

of '"P-labelled phosphate from an ai id lea soil of Sri Lanka, p. 345-3:36 In Isotopes and radi- 
ation in research on ii-plant relationsh In tomi n > n V i i ^ i u 

Smaling, E.M.A., S.M. Nandwa, and B.H. Janssen. 1997. Soil fertility in Africa Is at stake, p. 47-61. 
/ ' I ii I i ! i I il teitihty in Africa. SSSA Spec. Publ. 51. SSSA, 

Madison, WI (this publication). 

Smith, J.L., R.I. Papendick, D.F. Bezdicek, and J.M. Lynch. 1993. Soil organic matter dynamics and 
ii i idi ni i ig ,n ai | 65-94. In M.F. Blaine Jr. (ed.) Soil microbial ecology: 
Applications in agricultural and environmental management. Marcel Decker. New York. 
inn i I ' i I I mill 1980. I i i i i i n I 1 1 i i li phoru p. 515-544. 

i \\ tin in ( Dm u i 1 phorus in aaii ultu \ ml 

SSSA, Madison, WI. 


Staunton, S., and F. Leprince. 1996. Effect ofpH and some organic anions on the solubility of soil 
phosphate Impliiali i toil uoa nl biln Lurop. J. Soil Sci. 47:231-239. 

Stevenson. F..I 1 di I ' I i i u i u u \ ind quality of soil 

organic matter, p. 173-199. In D.C. Coleman et al. (ed.) Dynamics of soil organic matter in 

tropical ecosystems. Univ. of Hawaii Press, Honolulu. 
Strobel, H. (ed.). 1987. Fei til iz i t inn ndation jeci fin 1 eport Annex III. Description of 

the first priority sites in the various districts. Ministr. of Agric, Natl. Agric. Res. 

Laboratories, Nairobi, Kenya. 
Swift, M.J., P.G.H. Frost, B.M. Campbell, J.C. Hatton, and K.B. Wilson. 1989. Nitrogen cycling in 

farming systems derived from savanna: Perspectives and challenges, p. 63-76. In M. 

Clarkholin and L. Bergstrom (ed.) Ecology of arable land. Kluwer Academic PubL Dordrecht. 

the Netherlands. 
Swift, M.J., O.W. Heal, and J.M. Anderson. 1979. Decomposition in terrestrial ecosystems. Studies 

in ecology. Univ. of Calitoin Pi Berkele; 
mi i I [) i i i I Iugwh I Effective! i minimal area man i u f mitrienl 

foi n i i im I i ' in. I 81:31-35. 

Tian, G., B.T. Kang, and L. Brussaard. 1992. Effects of chemical composition on N, Ca, and Mg 
I i I n n incubation i I from lected a forestry i fall lant sped 

Biogeochemistry 16:103-119. 
i n i i I i ii I I i i I. 199 vlulchin fleet of plant residu vith chemically con- 

trasting compositions on maize growth and nutrients accumulation. Plant Soil 153:179-1X7 
Tusji G.Y., G. Uehara, and S. Balas (ed.). 1994. DSSAT V3. Univ. of Hawaii, Honolulu. 
Van der Kruijs, A.C.B.M., M.T.F. Wong, A.S.R. Juo, and A. Wild. 1988. Recovery of 15 N-labelled 

fertilizer in crops, drainage water and soil using monolith lysimeters in south-east Nigeria. J. 

Soil Sci. 39:483-492. 
Vaniauwe. B.. L. Dendooven. and R. Merckx. 1994. Residue fractionation anil decomposition: the sig- 
nificance of the active fraction. Plant Soil 158:263-274. 
Van Noordwijk, M., and S.C. Van de Geijn. 1996. Root, shoot ami soil parameters required for 

process-oriented mode is of crop growth limited b\ water or nutrients. Plant Soil 183: 1-25. 
mIi ill L. I inti i 1 ) ( i i i i in i ! ti i n | 1 i h ite and oxalate on an 

aluminum hydroxide montmorillonite complex. Soil Sci. Soc. Am. J. 55:65-70. 
Vogt, W, J. Sauerbom, and M. Honisch. 1991 Sin ,, ;. m< hi,, ii i d ution and infestation in 

Ghana and Togo on grain crops, p. 572-377. hi Proc. ot the 5th Ini. Synip. on Parasitic Weeds. 

Nairobi, Kenya. 24-30 June 1991. CIMMYT, Nairobi, Kenya. 
Whitmore, A.P., and E. Handayanto. 1997. Simulating die mineralization ofN from crop residues in 

relation \,- residue quality, p. ' 47-348. In G. Cadisch and K.E. Ciller (ed.) Driven b\ nature: 

Plant littei qualin and decomposition CAB lni.. Wallingford. England. 
Xu, Z.H., P.G. Saffigna, R.J.K. Myers, and AX. Chapman. 1993a. Nitrogen cycling in leucaena 

(Leucaena leucocephala) alley cropping in semi-arid tropics: I. Mineralization of nitrogen 

from leucaena residues Plant Soil 148:63-72. 
Xu, Z.H., P.G. Saffigna, R.J.K. Myers, and A.L. Chapman. 1993b. Nitrogen cycling in leucaena 

I ill || i in i i ii ] i II 1 i ii i lb to 

addition of nitrogen fertilizer and plant residues. Plant Soil 148:73-82. 

Gender and Soil Fertility 
in Africa 

Christina H. Gladwin, Ken L. Buhr, Abe Goldman, 
Clifton Hiebsch, Peter E. Hildebrand, Gerald Kidder, 
Max Langham, Donna Lee, Peter Nkedi-Kizza, 
and Deirdre Williams 

University of Florida 
( iainesville, Florida 


Population pressures, currency devaluations, and fertilizer subsidy removal pro- 
grams in many African countries have caused renewed concern about soil degradation and 
loss of soil fertility. Advances in food production are further constrained by the invisibili- 
ty factor, i.e., women do most of the food farming in sub-Saharan Africa, but have little 
access to the means necessary to significantly increase output and yields. We call this the 
invisibility factor because agricultural experts commonly do not acknowledge that most of 
Africa's smallholders are women, and women's yields, women's adoption, and women's 
use of inputs are rarely reported. Gendered differences in wealth result in women's low- 
ered access to cash and credit, needed to acquire both organic and inorganic fertilizers. The 
solution to this problem, we believe, lies in better collection of gender-desegregated data 
as well as better programs and policies that take into consideration the severe cash con- 
straints women farmeis face ind th tai i om n farmers with crucial inputs of produc- 
tion such as fertilizers. Because women lack cash, we recommend as a general objective 
achieving low application rates of about 25 kg nutrient ha" 1 , depending upon soil and cli- 
matic conditions. Options proposed to target women farmers with greater fertilizer inputs 
include fertilizer vouchers, providing fertilizer in small bags in local markets, microcred- 
it, free grants of fertilizer, use of organic materials, biological N : fixation technologies. 
combinations of organic and inorganic fertilizers, and improving women's access to cash- 
crop markets. As part of USAID's Soils Management Collaborative Research Support 
Program, we propose to test these different methods in several African contexts from 1997 
to 2002. 

There is a new sense of urgency about achieving food security for sub-Saharan 
Africa (hereafter referred to as Africa; Borlaug & Dowswell, 1995). Most African 
countries still depend on agriculture for most of their gross national product and 
employment (Tomich et al., 1995). Yet Africa currently imports a large propor- 
tion of its food grains, e.g., one-third of its rice (Oryza sativa L.) consumption 

Copyright © 1997 American Society of Agronomy and Soil Science Society of America, 677 S. 
Segoe Rd., Madison, WI 5371 1, USA. Replenishing Soil Fertility in Africa. SSSA Special Publicatii in 


and two-thirds of its wheat consumption (Eicher, 1995). Whether it can continue 
to do so is questionable. In 1995-1996, world commodity prices of major food 
crops skyrocketed. Because world stocks of rice, wheat {Triticum aestivum L.), 
maize (Zea mays L.), and other grains fell to their lowest levels in 20 yr, the world 
price of maize jumped from US$ 2.43 to more than US$ 5.00 a bushel; wheat 
prices also jumped from US$ 3.49 to more than US$ 6.00 in 1995-1996. 

Because Africa's population is expected to increase by 100 million during 
the next 6 to 7 yr, there is intense pressure to increase food supplies from domes- 
tic production (Eicher, 1995). Yet after 10 yr of structural adjustment programs 
that sharply devalued currencies and removed fertilizer subsidies, fertilizer prices 
are so high that fertilizer use on food crops is now often unaffordable in many 
African countries (Bumb et al., 1996). 

In most of Africa, advances in food production are further constrained by 
the invisibility factor, i.e., women do most of the food farming but have little 
access to the means necessary to significantly increase output and yields. On 
average, African women provide 46% of the labor inputs and are responsible for 
up to 80% of domestic food production in some societies (Dixon, 1982; Gladwin 
& McMillan, 1989). We call this the invisibility factor because agricultural 
experts commonly do not acknowledge that most of Africa's smallholders are 
women, and women's yields, women's adoption, and women's use of inputs are 
rarely reported. They correctly argue that development strategies need to reach 
African smallholders to be effective, but they ignore the fact that the constraints 
facing women smallholders may be an important part of the problem. Eicher 
(1982, 1995), for example, consistently does not mention that 45% of the small- 
holders responsible for Zimbabwe's second Green Revolution (1980-1986) are 
women; nor does he indicate the percentage of hybrid maize that was adopted by 
women or the percentage of fertilizer subsidies that went to women. Similarly, 
Smale's (1995) report on Malawi's delayed Green Revolution does not indicate 
women's adoption of hybrid maize; yet women's maize varieties, comprising 
95% of total maize production in 1987, are usually unfertilized local varieties 
consumed at home while men's hybrid varieties are sold as cash crops (Gladwin, 


The invisibility of women farmers has led to a debate about whether a turn- 
around in African agriculture can occur without helping women to farm. Most 
experts conclude that women farmers are essential for increasing Africa's food 
production, at least in the short run, because there arejust too many women farm- 
ing to ignore them (Dixon, 1982). Dixon's revision of International Labor 
Organization (ILO) estimates of women's participation in the labor force indicate 
that women represent on average 46% of the agricultural labor force in Africa 
when (i) subsistence production is counted as economic activity, (ii) unpaid fam- 
ily helpers are routinely included, (iii) all agricultural work is recorded, (iv) the 
survey is taken in the peak season, and (v) the definition of agricultural work 
includes gardening, raising poultry, and transporting crops to market. 


In the long run, however, African women farmers may be displaced from 
farming by men — just as black farmers in the southeastern USA were displaced 
by white farmers in the 1950s to 1970s (Gladwin & McMillan, 1989; Gladwin 
1996). Boserup (1970) suggests that intensification of agricultural production 
causes women's participation in farming to decrease relative to men's. Female 
farming systems are prevalent in African societies with low population densities 
and an ample land-to-person ratio such that families can produce their food with 
low labor inputs using shifting cultivation. These systems decline with population 
growth and agricultural intensification and are replaced by male farming systems 
as the plow is introduced (Boserup, 1970, p. 16-36). 

The displacement of women farmers with intensification, however, may be 
slowed down or mitigated by other location-specific factors (Gladwin & 
McMillan, 1989). These factors include extensive male outmigration from rural 
areas to the cities or mines for nonfarm work. In many parts of Kenya, female- 
managed households with migrant husbands account for 47% on average 
(Thomas-Slayter & Rocheleau, 1995, p. 14). In western Kenya, men often spend 
their lives working in the city, returning to the homestead a few times a year and 
retiring back to the countryside in their elder years. Other factors include the (i) 
resurgence of rinderpest and trypanosomiasis, which are slowing down the emer- 
gence of animal traction and thus the plow (FAO, 1983), and (ii) location-specif- 
ic soil types (i.e., sandy soils) and terrain (i.e., steep mountain slopes), which may 
make hand hoes preferable to plows even under permanent cultivation (Pingali et 
al., 1987). The intensification process and thus the displacement of women farm- 
ers is not likely to proceed at the same rate or in the same pattern in all African 

Yet women farmers have already been replaced in many parts of rural 
Africa, because development planning often still fails to include women, and new 
technology often continues to go only to the men farmers, despite all the concern 
with women in development (WID; Dey, 1981; Fortmann, 1981; Spring, 1986; 
Goheen, 1996). Most externally funded development projects are aimed primari- 
ly at men because they are run by men. The majority of extension agents are male, 
with a few exceptions housed in a poorly funded WID office (Staudt, 1975). Men 
tend to monopolize new capital inputs, whether they are mechanical or biologi- 
cal (e.g., plows or fertilizer), as regression results from data on 137 agricultural 
societies in the Human Relations Area Files show (Burton & White, 1984). Yet 
women farmers are interested in new technologies and want development inter- 
ventions (Due etal., 1983). In addition, numerous studies show that women farm- 
ers still lack access to (i) the basic agricultural inputs of land, labor, capital or 
credit, and organic and mineral fertilizers (hereafter referred to as inorganic fer- 
tilizers), (ii) extension advice, and (iii) the market and political arena (Staudt, 
1975;Elson, 1989; Due, 1991; Elabor-Idemudia, 1991; Goheen, 1991; Guyer& 
Idowu, 1991; Saito etal, 1994). 


Why haven't women been included in the past? Why has women's access 
d productive inputs been blocked? One rationale given is that men farmers are 


more productive than women farmers. It is true that the raw, unanalyzed data on 
yields of female-headed households (FHHs), which comprise 25 to 35% of 
African households, compared with those of male- or joint-headed households 
show that FHHs have less labor and plant smaller crop acreages, have less access 
to credit and plant more subsistence crops, and are therefore not as productive as 
men (Due, 1991; Due & Gladwin, 1991). An analysis of productive efficiencies, 
however, requires the proper estimation of a production function that controls for 
other explanatory variables besides gender, e.g., the studies of agricultural pro- 
ductivity done by Robert Evenson and his students (Bindlish & Evenson, 1993; 
Alderman etal., 1995; Quisumbing, 1996). When researchers estimate a produc- 
tion function that controls for other explanatory variables such as input levels 
(e.g., land, labor, capital, extension advice, and education), most studies show 
that male and female farmers are equally efficient as farm managers (Moock, 
1976; Bindlish & Evenson, 1993; Saito et al., 1994). ' When these other explana- 
tory variables are held constant while an independent gender variable is allowed 
to vary in a multiple regression analysis, researchers usually find that the inde- 
pendent gender variable (expressed as a dummy variable or intercept shifter) is 
not significant (Quisumbing, 1996). What this means is that "the gender yield dif- 
ferential is caused by the difference in the intensity with which measured inputs 
of labor, manure, and fertilizer are applied on plots controlled by men and women 
rather than by differences in the efficiency with which these inputs are used" by 
men and women (Alderman et al., 1995). Alderman et al. (1995) conclude that 
household output could be increased by 10 to 20% by reallocating the inputs (e.g., 
moving some fertilizer) from plots controlled by men to plots controlled by 

If women were given the same access to yield-increasing inputs as men, 
then the smallholder agricultural sector would see significant increases in agri- 
cultural productivity. African countries that address these gender disparities in 
input use and remove these barriers to women's productivity would increase their 
aggregate agricultural productivity. 


Why are women farmers not given the same access to yield-increasing 
inputs as men? The answer to this question requires an understanding of the spe- 
cial features of me African household. The African household is usually an 
extended rather than a nuclear family, with individual production and consump- 
tion units embedded within it. These units tend to be semi-autonomous and are 

i function would be estimated by ordinary least 

n T+ b In E + c EXT + d Gender + 

where Fis output; L is labor input (hired oi famih I: T 'is ;: vector of laud, capital, and other inputs; E 
is educational attainm ni L > T i an hid \ of extension rvi Gentler is th jender of the house- 
hold head or farm manager; and e is the error term. The coefficient that indicates gender differences 
is d, an intercept shifter (Quisumbing, 1996). 


often female-headed households headed by women such as the wife or wives (in 
polygamous societies) of the household head, or his daughters-in-law or sisters- 
in-law (in societies with substantial male outmigration of adult sons or younger 
brothers, who would normally live in the same rural compound with the house- 
hold head). The female-headed households may be de facto or dejure. A de facto 
female-headed household is one in which the husband is temporarily away, mak- 
ing it necessary for the wife to make at least some of die agricultural decisions 
and support the family, possibly aided by remittances from the husband. A dejure 
female-headed household is one in which the head is divorced, widowed, or a sin- 
gle parent and must make all decisions and provide all support for the family. The 
resource base of these two types of female-headed households could potentially 
be very different, depending on the size and frequency of the remittances. In an 
additional 14% of African societies (Vaughan 1986), the units may be headed by 
a woman simply because the society is matrilineal, one in which inheritance 
passes through the mother, or matrilocal, one in which children belong to and 
reside in their mother's, not father's, lineage or family. Autonomy of the unit 
comes from two sources. First, the woman in each unit has some responsibilities 
independent from the household head to feed or clothe or educate the children in 
her unit. Depending on the cultural rules, she may be responsible for certain foods 
or all the food during a certain period, e.g., the hunger months. Second, she ful- 
fills this responsibility with an income stream she herself generates independent- 
ly of the household head or her husband. 

These separate income streams are a second unique feature of the African 
household and have been well documented in me literature, e.g., Mossi women 
who own private fields in Burkina Faso (Gladwin & McMillan, 1989); husbands 
and wives who lend each other money at rates slightly less than the prevailing 
market rate; the payment of wages inside households; wives who sell water to 
husbands in the fields; husbands who sell firewood to wives; and both who sell 
animals to each other on festive occasions (Koenig, 1980; Guyer, 1981; Okali & 
Sumberg, 1986). In any of these exchanges, the best interests of me household 
may not coincide with those of particular members, so that it makes more sense 
to model me household as a collective firm — rather than a unitary entity — in 
which a wife's budget is delinked from her husband's, to varying degrees depend- 
ing on the particular culture, and wives respond to changes in their husbands' 
allocation decisions solely according to their own needs (Jones, 1983; Alderman 
etal., 1995). 

Usually, separate income streams give some autonomy to the women in the 
household, but they do not necessarily give power to the women heading up the 
unit, and this distinction is a third significant feature of the African household. 
Women are relatively powerless compared with the male household head (Moser, 
1989; Kabeer, 1994), who may interrupt women's work in their own fields and 
demand mat they supply labor to the cooperative fields he manages and from 
which he receives income (Gladwin & McMillan, 1989). Alternatively, he may 
demand that the wife goes to the store to buy fertilizer for him, or he may take 
her fertilizer to use on his fields or crops. Asymmetric power relationships with- 
in the African household, therefore, influence women's access to yield-increasing 
inputs of production as well as the fertility of their soils and the yields of their 


food crops. The question of who gets access to productive inputs is a political 
question — the result of a power negotiation — and not just an economic question 
(Bates, 1983). In a power negotiation, women in asymmetric power relationships 
often lose out to men with greater power, status, and prestige. 


Microlevel research in the 1980s has documented the constraints to 
women's use of inorganic fertilizers. Data collected in Malawi and Cameroon by 
Gladwin (1991, 1992) showed that the majority of African women farmers used 
no inorganic fertilizer because they had neither the cash nor the credit to acquire 
it. In Malawi, the average female-headed household used 34 kg ha" 1 of fertiliz- 
er — significantly less than the 51 kg ha- 1 of the male-headed household, and the 
median use of fertilizer by women was zero. Data collected from 36 households 
in anglophone and francophone Cameroon in 1989 agreed. Women's average fer- 
tilizer use was 30 kg ha" 1 compared with 52 kg ha" 1 for men, because two-thirds 
of the anglophone women farmers used no fertilizer at all. 

Are these gender differences because of gender itself or gender differences 
in wealth? The data from Malawi suggested that the lower resources controlled 
by women might better explain their lower use of fertilizer, because female-head- 
ed households owned only 0.8 ha of land compared with 1.33 ha owned by male- 
headed households (P = 0.0001; Gladwin, 1992). Regression analysis agreed. 
With the quantity of fertilizer used by a sample of 498 male- and female-headed 
households in 1986/87 as the dependent variable, results showed that member- 
ship in a credit club and use of manure/compost significantly increase the quan- 
tity of fertilizer applied per hectare (P = 0.0001 and 0.01); whereas variables of 
farm size and lack of cash significantly decrease the quantity of fertilizer applied 
per hectare (P = 0.0001). When these variables are included in the equation, the 
gender variable is not significant, but wealth variables are significant. 

Results of modeling smallholders' fertilizer decisions in West, East, and 
southern Africa also show the main reason women do not use inorganic fertilizer 
is their lack of cash, capital, or credit to acquire it, not their belief in organic fer- 
tilizers or a fear of dependency on inorganic fertilizers. All of these criteria were 
included in decision models to use both organic and inorganic fertilizer, or either 
or neither of them on maize in Malawi and Cameroon (Gladwin, 1991) and in 
Kenya (David, 1993; Williams, 1997). Among 75 farmers used to test the model 
in Malawi and Cameroon, 17 (12 of them women) eliminated both organic and 
inorganic fertilizers because of lack of cash or credit. Only 5 farmers did not use 
inorganic fertilizers because of the risk of their land's becoming dependent on 
inorganic fertilizer. 

Film igi iphii i b i rvations also showed that African women realize the need for inorganic fer- 
tilizer. In francophone Cameroon, for example, the) interplant their own food crops [e.g., maize and 
beans (Plia.s-eolu.s n 1 | in til i field thin n rops [ n t (Caj'fett ambica 

L.)]. They do this to siphon off some of the N fertilizer applied to men's crops to their own crops. 
After the crops are interplanted and while weeding then maize, women scrape off some of the N fer- 
tilizer still undissolved in the topsoil around the mens coffee, and push it nearer to their maize plants. 


What is the morale of this story? Just as in the productivity studies cited 
above, gender per se has no direct effect on fertilizer use. Although women 
household heads apply less fertilizer than men heads, gender does not matter 
when one holds constant the access to cash and credit. It is gendered differences 
in wealth and women's lower access to cash and credit that explain their lower 
fertilizer use. Without as much capital or credit, women apply less fertilizer than 
men — and get lower yields and incomes as a result. 


Women also face many constraints limiting their use of organic fertilizers. 
For example, women's lack of land constrains their use ofN2-fixing beans as a 
sole crop (Kumwenda et al., 1996) or their interplanting of Ni-fixing tree crops 
with maize. Women's lack of animals and pasture land limits their access to 
manure; and their lack of cash constrains them from buying it. While more than 
one-half (44 of 75) of the women farmers surveyed in Malawi and Cameroon 
believed organic fertilizer was needed on maize in addition to inorganic fertiliz- 
ers, almost one-half (20) of the 44 women did not use it because they lacked ani- 
mals and cash to provide the manure or compost (Gladwin, 1991). 

Hedgerow intercropping (HI), an agroforestry technique designed mainly 
to increase nutrient supply to annual crops, was tested in on-farm trials in Kenya 
by the International Centre for Research on Agroforestry (ICRAF), the Kenya 
Forestry Research Institute (KEFRI), and the Kenya Agricultural Research 
Institute (KARI) in the late 1980s and early 1990s. The research objectives 
included the selection of promising agroforestry tree species for HI, quantifica- 
tion of tree biomass production potential, and testing HI for its appropriateness to 
farmers. Leucaena [Leucaena leucocephala (Lam.) de Wit] and calliandra 
(Calliandra calothyrsus Meissner) tend to produce the highest biomass for use 
either as mulch or as fodder for livestock and thus have often been the recom- 
mended species for farmers. However, adoption of HI has been low (Swinkels & 
Franzel, 1997). Why? Williams (1997) used ethnographic interviews and deci- 
sion-tree modeling from a sample of 40 women farmers around Maseno in west- 
ern Kenya to elicit their reasons for adoption or nonadoption of HI technologies. 
She came to many of the same conclusions reached earlier by researchers of HI 
in western Kenya (David, 1992; Swinkels & Ndufa, 1993; Bekele, 1996; 
Shepherd et al., 1997; Swinkels & Franzel, 1997). As her decision-tree model in 
Fig. 9-1 shows, about one-third of the women had no knowledge of HI technol- 
ogy (Criterion 1). Other constraints included women's lack of access to seedlings 
(Criteria 2 and 3), lack of knowledge of how and where to plant them (Criterion 
4), and the belief that planting trees would actually lower soil fertility (Criterion 
6). More common, however, were shortage of land and labor (Criteria 8 and 10), 
especially in situations where agricultural intensity and population density were 
high. 3 Many female heads of households and women with small children felt that 
the large amount of labor required to coppice the trees would prevent them from 


using this technology. Where hedges were not frequently pruned, some farmers 
noted shading out of companion crops, which would result in reduced crop yields. 
Some farmers who had tried HI said that the trees were taking up "more room 
than the crops themselves" and were shading them out, or both; hence they decid- 
ed to uproot the trees. In addition, many women reported problems with pests 

{Employ HI with calliandra or leucaena; don't} 
40 cases 

1 <Do you know about W?> 
Yes | X. No 

<Were seedlings supplied by your women's N* Don't use I 

group, ICRAF, CARE, or another NGO?> I— 1 

Yes | V Mo "cases 

4 <Do you know how to plant Yes 3 <Afe gb|e tQ buy m acqujfe 
hedgerows among your crops?> seedlings from elsewhere?> 

Yes I \"° | No 

5 <Do you need or want more ^s| Don . tuse | [~Don't use | 

trees for fuelwood or fodder?> 

(1 error) 

.Use HI unless: | No ^^ | Don't use HI unless: | 

6 <Do you believe these trees <Do you have enough land 

could destroy your soil?> to plant hedgerows?> 

| Don't use | | No I | Don't use | 

cases s <Dq yQu have en0U g n 9 <Do you think these trees could 
land to plant hedgerows?> improve your soil fertility?> 

Yes | X 1 ^ Yes I I No 

10 <Do you have enough help 1 ° USe l I use HI I I Don't use I 

or can you afford hired 5 cases I 1 I 1 

laborers to cut back the trees < 1 error > 

when needed?> Nn\.i 1 

Yeg i ^ | Don't use | 4 cases 

11 <Do you think these trees could shade out 

your crops before you can cut them back or 

could they take up more room than the crops ?> 

Yes I [ No 

I Don't use I " <Do P ests and/or termites 

L— 1 attack these trees too much 

2 cases for tnem to be wortn g ro wing?> 

Yes{ [No 

| Don't use | | Use HI | 

2 cases 8 cases 
(2 errors) 

?ig. 9- 1 . Decision tree for the adoption or non-adoption of hedgerow intercropping (HI) by 40 w 
farmers from around Maseno in western Kenya (Williams, 1997). 


such as termites eating the seedlings, while a few experienced attack of leucaena 
by the psyllid pest (Heteropsylla cubana Crawford) and lacked the cash to buy 
pesticide. This resulted in a substantial decrease in biomass production, which 
caused women farmers to want to uproot the trees. The cumulative result of all 
these constraints was that the decision model predicts only 8 of 40 women would 
use HI technologies in this densely populated setting. These results complement 
those of Shepherd et al. (1997), who found little evidence that hedgerow inter- 
cropping improved yields under farmers' conditions. 


How can this lack of fertilizer use by women be turned around? To include 
women in the process of technology transfer, project planners should collect gen- 
der-desegregated data on yields and use of inputs, plant a proportionate percent- 
age of farmers' test plots on women's fields with women's food crops, and adopt 
policies specifically to target women farmers with crucial inputs of production 
such as fertilizer, all the while taking into consideration the severe cash con- 
straints women farmers face. These crucial constraints of lack of cash and credit 
determine the quantity and type of inorganic fertilizer women apply as much as 
do the specific crops women plant. 

Because women lack cash, we recommend as a general objective low appli- 
cation rates of about 25 kg nutrient ha" 1 or 50 kg fertilizer ha" 1 (Kumwenda et al., 
1996; Larson & Frisvold, 1996; Benson, 1997). Where beans or other legumes 
are the women's primary crop, low rates of P (11 kg P ha" 1 or less) may be 
required. In most African cropping systems where maize, other cereals, or tuber 
crops are women's major crops and are interplanted with various other crops, 
small amounts of high-analysis N fertilizer (e.g., 25 kg N ha' 1 ) can increase food 
production, because N is almost always the main limiting nutrient; however, low 
rates of soluble P fertilizer can effectively increase production of cereals on P- 
deficient soils. Jama et al. (1997), for example, found that 10 kg P ha" 1 as triple 
superphosphate (TSP) increased maize yield and was financially attractive on P- 
deficient acid soils in western Kenya. 

How can women farmers be specifically targeted? The development litera- 
ture describes eight general options, discussed below. 

Solutions Involving Use of Inorganic Fertilizer 

Fertilizer Voucher Option 

As a temporary measure, provide vouchers directly to cash-poor women 

farmers producing food crops in make it passible for them to obtain small 
amounts of fertilizer. A voucher system would allow an African government bur- 
dened with fiscal deficits to do something about food security by targeting the 
subsidy directly at women farmers who produce most of the food, and it would 
encourage healthy competition among private distributors in the fertilizer indus- 


try. With a voucher system, women farmers in women's clubs would receive 
vouchers to take to private fertilizer distributors, from whom they would buy fer- 
tilizer at a discount (similar to the way the poor in the USA buy food with food 
stamps or pay for housing with housing vouchers). The government would then 
remunerate distributors for the vouchers. The government's physical presence in 
the fertilizer distribution system would be minimized, and its total subsidy bill 
would be less than it was when fertilizer subsidies were freely extended to all 
growers of food and export crops, men and women alike. 

The vouchers would be discontinued after a number of years, and women 
would buy fertilizer from local merchants on the open market at the market price, 
with or without credit. The temporary program of vouchers would be coupled 
with a plan to supervise women's application of fertilizer, to reduce leakages — 
defined as the use of vouchers for other than women's crops. The plan would also 
strengthen the revolving credit funds used by many women's clubs to bail out 
individual defaulting members. Clubs would receive a stipend to supervise the 
application of vouchered fertilizer on women's fields. Women's clubs can thus 
serve not only to expand credit to women but also to supervise the proper use of 
fertilizer vouchers. 

Donors like the World Bank, however, have spent the last 10 yr removing 
fertilizer subsidies. Their policy now is to move toward full market cost of fertil- 
izers (Donovan, 1996; K. Saito, 1996, personal communication). In fact, most 
food policy analysts recommend that input subsidies, and particularly fertilizer 
subsidies be eliminated entirely, because they are a common technique used to 
increase the profitability of intensive agriculture while keeping food prices artifi- 
cially low. Timmer et al. (1983, p. 288) argue that "only when total fertilizer use 
is low and the ratio of incremental grain yield to fertilizer application is high" can 
such subsidies be cost effective, relative to higher output prices or greater food 
imports. African governments burdened with large fiscal deficits should therefore 
consider whether fertilizer subsidies represent die best use of their limited 
resources. After all, "someone must pay for the subsidy". Economists thus con- 
clude that "all subsidies tend to distort the intensity of use of inputs from tiieir 
economically optimal levels, and significant waste is a result. Since not all inputs 
can be equally subsidized, output price increases will have a greater impact on 
productivity than will input subsidies, especially in the long run" (Timmer et al., 
1983, p. 288). 

This line of reasoning makes sense when applied to Asia and Latin America 
now; but it did not make sense during their Green Revolution era in the 1960s and 
1970s, when fertilizer use contributed 50 to 75% of me increase in yields in food 
crops (Byerlee & Heisey, 1992), and adoption of fertilizer -responsive modern 
varieties depended on fertilizer subsidies (Harris, 1984; Van der Eng, 1994; 
Eicher, 1995; Goldman & Smith, 1995). Neither does it apply to current condi- 
tions in sub-Saharan Africa, where average fertilizer use — not nutrient use — is a 
mere 7 to 11 kg ha" 1 , and women food producers commonly use no fertilizer 
(Lele et al., 1989). Larson and Frisvold (1996) conclude mat average fertilizer 
application rates in Africa need to increase from 10 to 50 kg ha" 1 within 10 yr (an 
18% annual growth rate) to prevent mining of soil nutrients. Near-term environ- 
mental concerns in African agriculture stem more from the persistent decline of 


soil fertility rather than from overuse of fertilizers, since this leaves continuous 
expansion of cultivation to relatively unused areas as the only option for increas- 
ing total output. Policy interventions are thus needed to encourage women food 
producers to increase their yields of traditional as well as modern varieties, and 
fertilizer subsidies in the form of vouchers are the most direct policy tool plan- 
ners have at their disposal (Gladwin, 1991, 1992). From the viewpoint of the 
women farmers, such vouchers are preferable to an expansion of credit opportu- 
nities, because women face many constraints to credit use that men don't face: 
they are either too poor, too old, or lack control over a cash crop with which they 
can repay a fertilizer loan (Gladwin, 1992, 1996). The risk of borrowing is par- 
ticularly high for them because they may have to sell some of their subsistence 
crop in the hunger months when their children are hungry in order to repay die 
loan. Rather than take that risk, they will often decide not to get credit, not to use 
fertilizer, and not to increase their yields. 

Fertilizer subsidies can decrease diis risk for resource-poor women farm- 
ers, and so can play an important role in increasing their food production 
(Gladwin, 1997). For this reason, Eicher (1995) blames the donor community for 
failing to present a balanced view (e.g., in World Bank, 1994) of the substantial 
role subsidies played (and still play) in Asia's Green Revolution: "Currently 
donors in Africa are focused on a number of policy reforms such as correcting 
overvalued exchange rates and removing fertilizer subsidies rather than long- 
term, institution-building activities, the hallmark of donor assistance in Asia in 
the 1960s and 1970s. In their zeal to remove fertilizer subsidies in Africa, how- 
ever, some donors are neglecting to inform African policy makers about the role 
of subsidies in Asian agriculture." 4 

Pinstrup -Andersen (1992) claims that fertilizer subsidies can serve as a 
temporary measure to compensate for the factors that make it difficult for 
African, as opposed to Asian, entrepreneurs to freely compete in an open fertiliz- 
er market. Among these factors are (i) the small volume of fertilizer that most 
African countries import, which weakens their bargaining position in negotiating 
for lower prices; (ii) high transportation costs within most African countries; (hi) 
high storage costs, which increase the expense of fertilizer distribution; (iv) 
unpredictable government policies and unstable institutions, which scare off pri- 
vate entrepreneurs from investing in input distribution systems; (v) the relative 
ease of government's acquiring fertilizer in the past as foreign aid; and (vi) the 
tendency of governments to maintain large fertilizer stocks, which may be 
released anytime and at any price. Pinstrup-Andersen (1992) concludes that gov- 
ernments should privatize fertilizer distribution in a way that assures competition, 
or else the private sector fertilizer distribution system may be no more efficient 
than the public sector system it replaces. If monopoly profits accrue, it will be 
more expensive. He also believes that fertilizer prices can be brought down only 
if, in the long run, governments invest in the infrastructure to reduce transporta- 
tion and marketing costs; but until they do, "there is a place for fertilizer subsi- 
dies" to compensate for me factors resulting in very high fertilizer prices 
(Pinstrup-Andersen, 1992). 

:s are still widespread in Asia: e.g., Indonesia's implicit 


Small Bag Option 

Improve the availability of small amounts of fertilizer in local markets and 
shops by repackaging 50-kg bags of fertilizers into smaller bags. Traditionally, 
fertilizer has been sold in 50-kg bags. Since most fertilizer for family food pro- 
duction must be carried both to the home as well as to the food plots, the weight 
of the bag is an important issue, as well as the amount of cash or credit needed 
for the purchase. The transportation cost of fertilizer from the market to the home 
and the field also is a factor in its use. Having fertilizer available in smaller bags 
(complete with pictorial instructions) would make it more affordable for women 
and easier to carry. This strategy is compatible with the views of many econo- 
mists who believe that accessibility of fertilizer is the main constraint to its 
increased use (Lele et al, 1989). If fertilizer were widely sold in local markets 
like cement, and available in weights that could be headloaded home and to the 
fields, women farmers would be more likely to buy it. Also, small bags reduce 
the risk associated with open bags of fertilizer absorbing moisture and becoming 
difficult to use. For these reasons, the sale of fertilizers in 5-, 10-, and 20-kg bags 
at local markets should encourage women farmers to use more fertilizer. 

Microcredit Option 

Expand the fertilizer credit market for women farmers through community 
banks operating on the Grameen Bunk model. The Grameen Bank in Bangladesh 
targets very small loans to groups of virtually landless women producers (Von 
Pischke, 1991, p. 233; Khandker et al, 1995). With two million borrowers and a 
recovery rate of >90%, it is clearly a compelling model. By 1994, it served one- 
half of all villages in Bangladesh, lent about US$ 385 million, and mobilized 
another US$ 306 million as savings and deposits (Khandker et al, 1995, p. xi). 
The bank is unique in that its explicit goals are to alleviate poverty and create 
self -employment opportunities for illiterate people who own less than one-half an 
acre of land and have never received a loan from the formal financial system. 
Since 1985, it has specifically channeled credit to women, who are less empow- 
ered among the rural poor. Increasingly, women receive the bulk of the loans and 
are the majority of the members. Their share of total cumulative disbursement 
rose from a little more than one-half in 1985 to 9 1 % in 1994; while female mem- 
bership grew from 65.5% of the total in 1985 to >94% in 1994 (Khandker et al., 
1995, p. 25-26). Strict observance of the norms forces group members to be 
accountable to each other. The second two women receive their loans only if the 
first two of the five women in the group repay regularly, and the group leader is 
customarily the last to receive credit. This creates pressure among group mem- 
bers to enforce the contracts and helps screen out bad borrowers. Savings mobi- 
lization is thereby encouraged. In 1994 women's savings amounted to 74% of 
total savings mobilized (Khandker et al, 1995, p. 31). 

What lessons can Africa learn from the Grameen Bank? The first lesson is 
that a bank with poverty-alleviation goals can also be sustainable as a bank by 
lending at market interest rates. The Grameen's lending rate has been 20% since 
1991 (Khandker et al., 1995, p. 66), and its subsidy dependency index (SDI) has 
decreased over time from 180% in the 1980s to 36% in 1994 (Yaron, 1992, 1996). 


The second is that women are often better credit risks than men, since loan recov- 
ery rates for general loans have been higher for women (97% in 1992) than for 
men (89%) (Khandker et al., 1995, p. 18). Whether it can be replicated in Africa 
is now being tested by Sasakawa Global 2000 programs such as Benin's CREPs 
(Caisse Rurale d'Epargne et de Pret), which mobilize savings before loaning to 
farmers, 20% of whom are now women (Galiba, 1996). 

Free Bag Option 

Introduce, for a short time only, a system of grants of small bags of fertil- 
izer targeted at the poorest women farmers who may not know the value of fertil- 
izer or are not self-sufficient in food production. Then numbers may be substan- 
tial; Kumwenda et al. (1996) estimate they comprise 40% of the smallholder pop- 
ulation in Malawi. After a temporary period, this program would be phased out 
and replaced by local merchants selling small bags of fertilizer at the market 

Solutions Involving Organic Inputs 
and Biological Nitrogen Fixation 

Given women's constrained supply of cash, together with the removal of 
price subsidies on fertilizers and their rising costs, the majority of African women 
farmers may be compelled to rely only on organic sources of nutrients — espe- 
cially legumes that fix atmospheric N2 — as the only available strategy for 
increased soil fertility. At current levels of availability and use, however, organic 
inputs are rarely sufficient to meet crop demand for nutrients or maintain soil 
organic matter (Palm, 1995; Kumwenda et al., 1996). The use of inorganic fertil- 
izer might be supplemented or enhanced with use of organic sources of nutrients 
(Palm et al., 1997, this publication). The following are possible options for get- 
ting organic nutrients to women farmers. 

Organic Source Option 

Make organic materials of farm origin more accessible. In addition to serv- 
ing as sources of nutrients, organic materials can influence nutrient availability 
by (i) acting as an energy source for soil microbial activity, (ii) serving as pre- 
cursors to soil organic matter, (iii) influencing the release pattern of plant-avail- 
able nutrients, and (iv) reducing P sorption of soil (Palm et al., 1997, this publi- 
cation). In on-farm trials, options would include use of green manure, animal 
manures, improved fallowing, biomass transfer, and legumes as sole crops in 
rotation or as intercropped with cereals (Giller et al, 1997, this publication). 
Information can be disseminated through extension workshops and field days for 
women and gender training of trainers for extension agents. Microcredit pro- 
grams can be used to improve access to organic inputs for women farmers. 

Biological Nitrogen Fixation Option 

Make biological N 2 fixation technologies more accessible. Such N2 -fixing 
crops as velvetbean [Mucunapruriens van utilis], pigeonpea [Cajanus cajan (L.) 


Millsp.], sunnhemp fCrotalaria juncea L.), lablab bean (Lablab purpureas), and 
crotalaria Crotalaria ochroleuca G. Don and agro forestry technologies with In- 
fixing trees could be promoted by making seeds, small loans, and extension edu- 
cation more accessible, and by devoting women farmers' test plots to intercrop- 
ping or rotation of legumes with cereals. Giller et al. (1994) conclude that N2 fix- 
ation from legumes can sustain tropical agriculture at moderate levels of output. 
The applicability of legume seed inoculation with rhizobia bacteria to con- 
ditions faced by African women farmers needs to be further tested. Although pre- 
vious studies show legume inoculation is simple, inexpensive, and highly suc- 
cessful in increasing crop yield (Meisner & Gross, 1980), the African experience 
is that this invaluable technology is largely unavailable to women farmers who 
need it most (Hubbell, 1995). In Uganda, Dr. Mary Silver of Mekerere University 
has been using NifTal inocula strains on soybean [Glycine max (L.) Merr.] as a 
test crop. She is testing them with both women and men farmers to see if non- 
adoption is due to lack of knowledge about biological N? fixation or rhizobium 
inoculants, lack of access to rhizobia when needed, lack of knowledge on how to 
use them, lack of seed, poor soils, agronomic characteristics of the crop, incon- 
sistency of extension advice, or to other infrastructural constraints such as the 
ineffectiveness of coating material such as gum arabic, syrup, and molasses, 
which attract hordes of ants that eat the seeds (C. Wortmann, G. Elkan, 1996, per- 
sonal communication). 

Organic-Inorganic Option 

Make combinations of organic and inorganic inputs in small amounts more 
accessible. Organic materials are frequently in limited supply and hence cannot 
by themselves provide the productivity boost needed by smallholders. The com- 
bination of available organic materials with the judicious use of inorganic fertil- 
izers may be a very appropriate option for women smallholders (Palm et al., 
1997, this publication). 

Cash Crop Option 

Introduce a cash crop into women's subsistence fanning systems. 
Sustainable food production is an important goal of development, and only when 
women farmers have cash will they have a sustainable way either to buy cash 
inputs or to repay loans for organic or inorganic fertilizers. In Malawi, women 
farmers are now growing burley tobacco (Nicotiana tabacum L.) and using its 
receipts to pay back loans for fertilizer use on both subsistence maize and tobac- 
co (Brown etal., 1996; Stephen Carr, 1996, personal communication). 


Given these diverse ways to target women farmers, which one should a 
rural development project use? As part of USAID's Soils Management CRSP 
(Collaborative Research Support Program), we at the University of Florida pro- 
pose to test these different methods in several African contexts during the 5 yr 


from 1997 to 2002, using three evaluation criteria. First, do women fanners, once 
exposed to one or more of these methods to increase fertilizer inputs, continue to 
use them? The women themselves are the only real experts on what method 
works for them and whether their yields have increased as a result. Second, how 
much leakage from women's to men's use is there with each method? Because 
women lack power and money within the African household relative to men, they 
may be pressured to apply their fertilizer to men's export crops, not their own 
food crops. Third, how sustainable over time is each method relative to the other 
options? Given that sustainable food production and food security are our prima- 
ry goals, it is necessary to see if and how the food producers themselves, most of 
whom in Africa are women, manage the fertility of their soils in the long run. 


The authors are collaborators in the University of Florida's Soils CRSP 
(Collaborative Research Support Program) funded by US AID. We are grateful to 
Steven Breth for editorial suggestions and Anne Thomson, Sylvia Coleman, and 
Charles Sloger for theoretical suggestions, two anonymous reviewers, and Steve 
Franzel and Roland Buresh for editorial c 


Alderman, H., J. Hoddinott, L. Haddad, and C. Udry. 1995. Gender differentials in farm productivi- 

t\ : Ii pli n a i hi lold efficien nd i ulti nil poli I I isi mpti n u uti 
Div. Dis. Pap. 6. Int. Food Policy Res. Inst., Washington, DC. 

Bates, R. 1983. Markets and states. Univ. of California Press, Berkeley. 

I ii s in western Kenya. 

M.Sc. thesis. Dep it'Foresti vedish I ni\ o1 siri ultural in es. Uppsala. 

Benson, T. 1997. The 1995/96 fertilizer verification trial— Malawi: Economic analysis of results for 
policy discussion. Rep. by Action Group I. Maize Producth it) Task Force. Ministry of Agric. 
and Livestock Dev., Gov. of Malawi, Lilongwe. 

Bindlish, V., and R. Evenson. 1993. Evaluation of the performance of T&V extension in Kenya. 
World Bank Tech. Pap. 208, Africa Tech. Dep. Ser. World Bank, Washington, DC. 

Borlaug, N.E., and C.R. Dowswell. 1995 ITi import nee o1 i iili 1 i ind i table adequate food 
supply to the II I m f liuinai I. p. I I I I i I.I I seaich in domestic 

and international agribusiness management Vol. I I. J A I Press. (Greenwich, CT. 

Boserup. E. I l )7<!. Woman's rule in economic de\ elopnient. St. Martin's Press, New York. 

I n I i I ii I i in n I Mi mi i i i n Let oriented econ- 

omy. Res. Rep. 1011. Agric. Policy Analysis Project Phase III. USAID Contract LAG-4201- 
Q-16-3061-00. Lilongwe, Malawi. 

Bumb, B.L., J.F. Teboh, J.K. Atta, and W.K. Asenso-Okeyre. 1996. Policy environment and fertilizer 
sector development in Ghana. In Proc. of the Natl. Workshop on Soil Fertility Management 
Action PI in fo ' Ii in i Efficit ill i il i mi manag in ill hallen for tht I si nm 

ry, Cape Coast, Ghana. 2-5 July 1996. Sasakawa Global 2000, Mexico City. 

Burton, M., and D. White. 1984. Sexual division of labor in agriculture. Am. Anthropol. 86:568-83. 

Byerlee, D., and P.W. Heisey. 1992 Stiat ies for I nical ch n n II i mi agncultuie with 

particular reference to sub-Saharan Africa, p. 21-52. In N. Russell and C. Dowswell (ed.) 
Policy options fur agricultural development in sub-Saharan Africa Ctr. for Appl. Stud, in Int. 
Negotiations (CASIN)/Sasakawa Global 2000, Airlie House, VA. 
1 nl 199 Open th d i u ind - ill th p o I Inti i I u In I. i i and the adoption of 

liedtierow intercropping. In Presented at the Rockefeller Foundation Social Science Meeting. 
CIMMYT, Mexico. 9- 1 3 November. CIMMYT, Mexico, D.F. 


David. S 1993. Manure or fertilizer: Decision-making in agriculture anions farmers i 

Kenya. ICRAF, Nairobi, Kenya. 
Dey, J. 1981. Gambian women: Unequal partners in ric 

Nelson (ed.) African women in the development pi 
Dixon, R. 1982. \\ I I I untries. Popul. Dev. 

Rev. 8:558-559. 
Donovan, W.G 1996. Agricull i in nom reform in sul haran Africa. Africa Tech. Dep. 

Environ. Sustain. Dev. Div. Working Pap. 18. World Bank, Washington, DC. 
Due, J.M. I I i i t ii ill mi nt pi i mi n 

African tem il Ids. p. 10 3-127. In C. Gl I 1 d 

African women farmers. Univ. of Florida [less. Gainesville. 
Due. .1.. and " Gladwin 19 I Inn ct 1 tri rural Iji tiiienl >i s ill on ifrican omen farm 

ers and female-headed households. Am. .1. Agric. Econ. 73 1431 — 1439. 
• i I I i 1 i 1 nd I nil I How in il nen percei i I nment 1 A case study 

n Zambi [ t ! i Dep. . * i ultui !ii m mi. (ii il llini ii In in 

Eicher, C.K. 1982. Facing up to Africa's food crisis. Foreign Affairs 61:151-174. 

Ii I mi I mi i ii i in ii i ditions foi lephcation. World 

Dev. 23:805-818. 
Elson, D. 1989. The impact of structural adjustment on women: Concepts and issues, p. 56-74. fa B. 

Onimode (ed.l Tl 1MI th World I nk uid th fri n deb il _ I Book London 

Elabor-Idemudia, P. 1991 The imj tofsti u Ijustm i i ram i men and their house- 

10I1 nl ml 1 in. i . un ii in 1 130. In C. Gladwin (ed.) ru tui ! dju i 

ment and African w m i in Un Florid t'i Gainesville. 

FAO. I In i i i mil i 10 ' ill i in i i >eual Supple. Issue. 

Food and Agric. Org, in. of the United Nations, Rome. 
1 I 1 I I I II f ncultutal policy on 

women, p. 205-214. In R. Dauber and M. Cain (ed.) Women and technological change in 

developing countries. Westview, Boulder, CO. 
Galiba, M. 1996. Rural nuance at smallholders' level. Sasakawa Global 2000. Benin-Togo, Cotonou, 

Giller, K.E., G. Cadisch, C. Ehaliotis, E. Adams, W.D. Sakala, and PL. Mafongoya. 1997. Building 

soil nitrogen ipital in Africa, p. 1 ;■> 1 - 192. /// R..I. Buresi t al. led.) 1 ileni shins i I fertil- 
ity in Africa. SSSA Spec. Publ. 51. SSSA, Madison, WI (this publication). 

Gill If I.F i Don sh md < Cadi i 1994 Can bio i nitrog i i ition sustain agricul- 
ture in the tropics? p. 1 73- 1 9 1 . In J.K. Syers and D.L. Rimmer (ed.) Soil science and sustain- 
able land management in the tropics, CAB Int.. Wallingford. England. 

Gladwin. C.H. I I rtili r sul . i I j i u ind theii tential impaci n m n i n 

ers in Malawi and Cameroon, p. 191-216. In C Gladwin led.! structural adjustment and 
African women farmers. Univ. of Florida Press. Gainesville. 

Gladwin. C.H. 1992. Gendered impacts of fertilizer subsith removal programs in Malawi and 
Cameroon. Agric. Econ. 7:141-153. 

Gladwin, C.H. 1996. Gender in research design: Old debates ai 
(ed ) Achiev in si ! i mp ict from i u i in . tmei 
Mexico City, Mexico. 

Gladwin, C.H. 1997. Targeting women farmers to increase food production in Africa, p. 55-71. fa S. 
i I I 1 I II ii 1 i n i t u hou 111 i I \ 

.Assoc. Mexico City. Mexico 

Gladwin. C.H.. and D. McMillan. 1989. Is a turnaround in Africa possible without helping African 
women to farm? Econ. Dev. Cultural Change 37:279-316. 

Goheen, M. 1991. The ideology and political economj of gender: Women and land in Nso, 
Cameroon, j 9 In i ladwin tructural adju tm it and Al i men 1 on- 

ers. Univ. of Florida Press, Gainesville. 

Goheen, M. 1996. Men own the fields, women own the crops. Univ. of Wisconsin Press, Madison. 

Goldman, A., and J. Smith. 1995. Agricultural transformations in India and Northern Nigeria: 
Exploring the nature of Green Revolutions. World Dev. 23(2):243-263. 

Guyer, J.I. 1981. Household and community in African studies. Air. Stud. Rev. 24(2): 87-137. 

Gu .11 with O Idowu 1991 •• men sricultural > . in multimodal rural economy: Ibarapa 
District, Oyo State, Nigeria, p. 257-280. fa C. Gladwin (ed.) Structural adjustment and 
African women farmers. Univ. of Florida Press. Gainesville. 

Harris, GT. 1984. Fertilizer subsidies in developing countries. Spec. Publ. SP-3. Int. Fertilizer Dev. 
Ctr.. Muscle Shoals, AL. 


nitrogen fixation technology in developing 

Jama, B., R. Swinkels, and R.J. Buresh. 1997. Agronomic anil economic evaluation of organic and 

inorganic sources of phosphorus in western Kenya. Agron. J. 89:597-604. 
Jones, C. 1983. The mobilization of women's labor for cash crop production: A game theoretic 

approach. Am. J. Agric. Econ. 65:1049-1054. 
Kabeer, N. 1994. Reversed realities: Gender hierarchies in development thought. Verso, New York. 
Khandker, S.R., B. Khalily, and Z. Khan. 1995. Grameen Bank: Performance and sustainability. 

World Bank, Washington, DC. 
Koeni L 19X0 [ousel Id behavior in the i on of Kita and it i 1 ti< n hi] to igri ultu I m :< 

p. 75-92. /nProc. of the 2nd Wot ksli t n leliai ,1 lii n DC. 20-24 Nov. 

Dep. ot Agricultural Economics. Purdue Univ.. West Lafayette. IN. 
Kumwenda, J.T., S.R. Waddington, S.S. Snapp, R.B. Jones, and M.J. Blackie. 1996. Soil fertility man- 
in ii h for tl ii iz iii iii i m II i I 1 i in southern Africa: A 

review. Nat. Resour. Group Pap. 96-2. CIMMYT, Mexico, D.F. 
L i i ii 1 i I I f I i i ] I i i i I n I I i m i i 

Africa: What is needed and why. Food Policy 21(6):509-525. 
Lele. L'.. R.E. Christiansen, and K. Kadiresan 1989. Fertilizer policy in Africa: Lessons from devel- 
opment programs and adjustment lending, 1970-87. Managing Agric. Dev. in Africa 

(MADIA) Dis. Pap. 5. World Bank, Washington, DC. 
lei sner. < a and H.D i 1980 lom gui I hit for tin valuation < th n d foi nd i pen 

i in ill n i f tropical legunn [ h. Bull > irtli Can Im i Agric. I Raleigh 

Moock, P. 1976. The efficiency of women as farm managers: Kenya. Am. J. Agric. Econ. 58:831-835. 
i i O. 19 i I i i i hi i ll ] u \ Id Ieeting practical ai i te i endei needs. 

World Dev. 17(11):1799-1825. 
Okali, C, and J.E. Sumberg. 1986. Sheep and goats, men and women: Household relations and small 

ruminant produ in n n outhw i ii ria p. 166-181. In J. Moock (ed.) Understanding 

Africa's rural house Ids a tannin \ i ii i ukler. CO 

hi 1)' in H in hi i i i i n >i i Mi in m i juiremen t inteiuopped plants. 

Agrofor. Syst. 30:105-124. 
Palm, C.A., R.J.K. Myers, and S.M. Nandwa. 1997. Combined use of organic and inorganic nutrient 

sources for soil fertilil m nan u in m | I 1 M7 /nR. J. Buresh etal. (ed.) 

Replenishing soil fertility in Africa. SSSA Spec. Publ. 51. SSSA, Madison, WI (this publica- 

Pingali. P.. V. Bigot, and H. Binswanger. 1 9S7. Agricultural mechanization and the evolution of farm- 
ing systems in ill n n i h i i In H ol in I in Pi I u in i 1 i 

Pinstrup-Andersen. P. 1993 Fertilizer subsidies: Balancing short-term responses with long-term 
imperatives, p. 93-99. In N. Russell and C. Dowswell (ed.) Policy options for agricultural 
development in sub-Saharan Africa. Ctr. for Applied Studies in Int. Negotiations 
(C ASlXi/Sasakawa Global 2000. Airlie House, VA. 

Quisumbing, A.R. 1996. Male-female differences in agricultural productivity: Methodological issues 
and empirical - id n i Ei in I >i \ ' ultui I h ng( J I '79-1596. 

Saito. K.. ii. Mekonnen. ami L). Spurling. 1994. Raising the productivity of women farmers in sub- 
Saharan Africa. Dis. Pap. 230. World Bank, Washington, DC. 

Shepherd, K.D., J.K. Ndufa, E. Ohlsson, H. Sjogren, and R. Swinkels. 1997. Adoption potential of 
ledgenn i in rcropping in mai ppii terns in the highlands o1 rn Ken 

I I ' i in i i i n i ii In n i >i I i 

Smale, M. 1995. Maize is life: Malawi's delayed Green Revolution. World Dev. 23:819-831. 

Spring. A. 1986. Men anil women participants in a stallfeeder livestock program in Malawi. Human 
Organ. 45(1): 154-162. 

Standi. K. 1975 Women Farmers and inequities in agricultural services. Rural Afr. 29:81-93. 

Swinkels, R., and '■hull' yd iptii n poteni I ! < I • i » inten oping systems in the 
highlands of western Kenya: II. Economic ami farmers' evaluation. Exp. Agric. 33:21 1-333. 

Swinkels, R., and J.K. Ndufa. 1993. Biol i i 1 in mi no i ,nn i sment of hedgerow 

1 r I I i 1 i i i I 

Central Africa AFRENA Workshop, Kabale, Uganda. 6-10 Sept. ICRAF, Nairobi. 

Thomas-Slayter, B., and D. Rocheleau. 1995. Gender, environment, and development in Kenya: A 
grassroots perspective. Lynne Rienner Publ., Boulder, CO. 

Timmer, C.P., W.P. Falcon, and S. Pearson. 1983. Food policy analysis. Johns Hopkins Press, 
Baltimore, MD. 


Toniich. T.P.. P. Kilby. and B.F. lolinston. I')95. Transforming agrarian economies Cornell l"ni\ 

Press, Ithaca, NY. 
Van cler Eng. P. 1994. Development of seed-fertilizer technology in Indonesian nee agriculture. Agric. 

History 68(l):20-53. 
Vaughan, J.H. 1986. Population and social organization, p. 1 59- 1 80. fa P. Martin and P. O'Meara (ed.) 

Africa. Indiana Univ. Press. Bloomington. 
Von Pischke, J.D. 1991. Finance at the frontier. World Bank, Washington, DC. 
Williams. D.E. 1997. (lender and integrated resource management: The ease ot western Kenya. M.A. 

thesis. Univ. of Florida, Gainesville. 
World Bank. 1994. Structural adjustment in Africa. Oxford Univ. Press, New York. 

ill i I i ii i i i li lie i in ill jl> JL 

Dis. Pap. 174. World Bank, Washington, DC. 
ron I 1' o Perfi mai e o1 > lopm n finano instituti n How to assess it? Organ, for Econ. 

Coop, and Dev., Paris. 

Ecological Economics of Investing 
in Natural Resource Capital 
in Africa 

Anne-Marie N. Izac 

Internationa] Centre for Research 
Nairobi, Kenya 


An ecological-economic analysis of soil fertility replenishment in Africa is pro- 
posed. Concepts from ecological economics and thermodynamics show that P and N are 
two different kinds of natural capital. Phosphorus is nonrenewable, whereas N is renew- 
able. The human use of non-renewablt I mrabi i ul in inti rg :n< rational externali- 
ties. The stock of P to which unborn generations will have access will be depleted as cur- 
rent and past uses transform a low-entropy resource into a higher entropy one. The basic 
economic questions are how best to liquidate nonrenewable P inventories and how best to 
invest in renewable N. A proposed rule stipulates that a socially optimal way of liquidat- 
ing P inventories and investing in N is to invest in soil organic matter — renewable natur- 
al capital with low entropy — by an amount equal to the costs of depleting P, which 
includes losses in ecosystem and agricultural amenities. Such a solution is superior to sole 
use of manufactured fertilizers (high-entropy resource) as substitutes for P and N because 
it ensures the maintenance of stocks of low-entropy natural capital. The implementation of 
this rule necessitates applications of phosphate rock (PR) together with appropriate quan- 
tities of organic inputs, in conjunction with soil conservation measures and, in some cases, 
applications of mineral N fertilizer. Given the positive environmental and social external- 
ities associated with these interventions, an equitable mechanism would be to share the 
costs of implementation among the different groups in society that benefit. This principle 
of the beneficiary pays or cost- sharing is used in European and North American agricul- 

Other authors that contributed to this publication document the extent of soil 
nutrient withdrawals occurring in African agriculture (Sanchez et al., 1997; 
Smaling et al., 1997). Even though the baseline data are not available to enable 
us to determine what kinds of trends in nutrient balances have taken place over 
time, there is a broad consensus (Stoorvogel et al., 1993; Rhodes et al., 1996) that 
current nutrient balances are negative in most of sub-Saharan Africa (hereafter 
referred to as Africa). There is also some evidence that these balances may have 

Copyright © 1997 American Society of Agronomy and Soil Science Society of America, 677 S. 
Segoe Rd, Madison, WI 5371 1, USA. Replenishing Soil Fertility in Africa. SSSA Special Publication 

been negative for a number of decades in some parts of Africa (Bruno, 1930). 
Nutrient withdrawal is highest in the relatively fertile and steeper parts of the con- 
tinent, viz., eastern and southern Africa (Smaling et al., 1997, this publication). 
Projections indicate that under current conditions, the extent of net withdrawals 
will increase in the foreseeable future throughout Africa (Stoorvogel et al., 1993). 
This suggests that nutrient pools will be decreasing over the coming years. 
Whether these decreases reach a threshold level of irreversibility and whether 
such a threshold even exists cannot be determined on the basis of currently avail- 
able data. It is nevertheless clear that processes of nutrient depletion in African 
agriculture are very significant and have serious implications for future agricul- 
tural production in the continent. It also is clear that N and P are the two most 
limiting nutrients in African agriculture (Bekunda et al, 1997). 

Economic analyses of soil nutrient use and withdrawals and soil nutrient 
management processes are sparse. There are nevertheless some key economic 
concepts, such as natural capital, that can help illuminate the debate about soil 
nutrient depletion from a different perspective than presented by other chapters in 
this publication. 

Economists have long been interested in evaluating the worth of the capi- 
tal of a nation (e.g., Adam Smith on The Wealth of Nations in the 19th century). 
The role played by nature in the formation of this capital has also been a topic 
much discussed since the Physiocrats — one of the first economic schools of 
thought initiated in France in the 18th century — argued that agriculture was the 
only real source of economic surplus in a nation, because it is founded on the free 
gifts of nature. Later on, Ricardo, an English economist from the 19th century, 
spoke of the original and indestructible powers of the soil. There has indeed been 
a long tradition in economics of considering that nature is boundless, and there- 
fore, by definition in economics, a free good. Consequently, agricultural produc- 
tion, which is the process of transforming nature into agricultural goods, has tra- 
ditionally been considered to be limited only by technical progress and certainly 
not by the resource base of agricultural systems. One of the basic assumptions of 
traditional or mainstream economic theory is indeed that of total substitutability 
of all inputs for one another, including nature, assumed to be totally substitutable 
by capital in agricultural production. 

It is only since about 1970 that a few economists have started to point out 
that nature's productive capacity for fulfilling human needs is actually not bound- 
less and that the purposeful manipulation of ecosystems in order to obtain agroe- 
cosystems with high rates of production or output entailed a number of environ- 
mental costs, and maybe even some irreversible losses (e.g., Georgescu-Roegen, 
1971). Even more recently, since about 1990, the concept of natural capital was 
developed, or rather resurrected, in the context of national (country) accounts. 
The concept has been used to measure a country's actual net income by taking 
into consideration the fact that natural capital depletion (e.g., mining of mineral 
deposits) occurs in many countries. In most national accounts the consumption 
and use of natural capital, such as mining minerals, is still considered as a source 
of income without the corresponding decline in the stock of the resource being 
accounted for as natural asset depreciation. Only manufactured capital is depre- 
ciated in such accounts. Finally, the concept of natural capital has also been used 


since about 1994 by a few economists interested in identifying the optimal level 
of natural capital necessary to ensure sustainability in a country (Whitby & 
Adger, 1996). 

These relatively recent attempts at explicitly recognizing the limits that 
scarce natural resources place upon agricultural production fall within the branch 
of economics known as ecological economics. In this chapter the issue of nutri- 
ent depletion in African agriculture is analyzed within the framework of ecolog- 
ical economics, which differs substantially from mainstream economics. The 
emerging field of ecological economics addresses the relations between ecosys- 
tems and social and economic systems. Social and economic systems are viewed 
as subsystems of the biosphere and thus as wholly dependent upon 
ecological-economic interactions. This is in direct contrast with the convention- 
al or mainstream economic approach, which considers that all phenomena are 
subsumed within and obey the rules of an economic system. When discussing the 
role of technical progress in alleviating resource constraints, the mainstream 
approach is perhaps best exemplified by the following statement from economics 
Nobel laureate Robert Solow (1974): "The world can, in effect, get along with- 
out natural resources, so exhaustion is just an event, not a catastrophe." 

The objectives in this chapter are (i) to analyze the process of nutrient 
depletion in Africa and the proposed soil replenishment initiative from an eco- 
logical-economics perspective, (ii) to determine what the policy implications are 
of this analysis, and (iii) to identify those issues that necessitate further investi- 
gation before a successful (from an ecological-economics perspective) soil 
replenishment project as proposed by Sanchez et al. (1996, 1997, this publica- 
tion) can actually be implemented in Africa. 


Natural capital is defined here as stocks of resources generated by natural 
biogeochemical processes and solar energy that yield flows of useful services and 
amenities into the future (Daly, 1994). Thermodynamics, a branch of physics 
dealing with conservation and changes of energy in systems, can help to provide 
insights into the physical characteristics of natural capital. 

In thermodynamic terms, natural capital is made up of low entropy 
resources (Georgescu-Roegen, 1971). The concept of entropy refers to the level 
of available energy or potential to be converted into work — as defined in physics, 
not economics! — of a system or a resource. A low entropy implies a high poten- 
tial to be converted into work. Energy cannot be created or destroyed (first law of 
thermodynamics), but it changes qualitatively from a state in which it is readily 
available and is of high quality (low entropy) to a high entropy state where it is 
bound, that is, where it is of low quality and is not readily available for further 
transformation (second law of thermodynamics). An implication of the second 
law of thermodynamics is that entropic degradation (moving from a low towards 
a high entropy state) characterizes all production processes, including agricultur- 
al production under conditions of intensification when more and more mechani- 
cal and chemical inputs are used to replace and boost natural capital inputs that 

produce too low an output to satisfy human needs (Georgescu-Roegen, 1971, p. 

The above definition indicates that soil nutrients are part of natural capital. 
Their pools are stocks of natural capital and by virtue of their energy content they 
contribute to a number of flows of ecosystem and agricultural amenities and ser- 
vices. These flows and services include the nutrient cycles, soil fertility, and plant 
nutrition that in turn contribute to agroecosystem productivity, resilience, stabili- 
ty, and sustainability. 

Natural capital can be either renewable (e.g., fish, water) or nonrenewable 
(e.g., minerals). The measuring rod of renewability is anthropocentric; it is the 
length of time scale needed for the natural reproduction of the stock of natural 
capital. Renewable natural capital can be naturally reproduced within a human 
time scale, through solar energy (Cleveland, 1994). Nonrenewable natural capi- 
tal, in contrast, is naturally reproduced at time scales that go far beyond the inter- 
est of humans (Cleveland, 1994). The distinction between renewable and non- 
renewable natural capital thus hinges on the different time scales during which 
natural (in contrast to human-induced) reproduction occurs. 

The two nutrients considered in this chapter fall into each one of these cat- 
egories. Nitrogen is renewable natural capital because its pool or stock can be 
increased through atmospheric biological N2 fixation within a relatively short 
period of time (e.g., one season; see Giller et al., 1997, this publication). 
Likewise, soil organic matter (of which N is a component) is renewable natural 
capital because it is regenerated through photosynthesis and plant death and 
decomposition over relatively short spans. Phosphorus, however, is nonrenew- 
able natural capital as it is regenerated by geochemical processes, which occur 
over periods of centuries to millennia. The different soil pools of P have different 
time scales or degrees of nonrenewability. See Buresh et al. (1997, this publica- 
tion) for a discussion of these different pools. 

The above definition of nonrenewable natural capital, which is based on 
both economic and thermodynamic concepts, implies that the continued use of 
such capital can only result in a decrease over time in stocks and flows of this 
capital and thereby in intergenerational externalities. African farmers, for exam- 
ple, have been reducing (mining) the stock of P in their soils over time (through 
crop harvests, erosion, and runoff) to the point where future generations of farm- 
ers will inherit P-deficient soils with lower productive capacity and reduced 
ecosystem functions. In economic terms, the action of one generation of farmers 
is negatively affecting the welfare of future generations of farmers. This negative 
externality raises issues of equity of access to nonrenewable natural capital 
among generations (intergenerational equity), which are discussed in the next 

The use of renewable natural capital also can trigger intergenerational 
externalities, if the rate of use or harvest of a resource is greater that its rate of 
regeneration. Thus pools of N can be depleted or even exhausted if the rate of N 
uptake and withdrawals through harvest is higher than the rate at which N is 

This issue is illustrated in Fig. 10 — 1, which shows that a shift in the pro- 
duction function of the agricultural sector occurs over time (from fi to f 3 ) as nutri- 


ent capital decreases and additional inputs are needed to maintain yields at the 
same level as before. This situation can be compared with that in southeast Asian 
and temperate agriculture where agricultural intensification has meant that a mar- 
ginal substitution of manufactured capital (defined as manufactured and cultural 
capital and as the transformation of natural capital into goods and services) for 
natural capital has occurred over time. This substitution has been partial in the 
sense that, contrary to Solow's contention, manufactured capital cannot be a total 
substitute for natural capital, as it requires natural capital as an input into its trans- 
formation process (Daly, 1994). In addition, this partial substitution has resulted 
in what in thermodynamics is called entropic degradation or dissipation of ener- 
gy, as low-entropy resources (e.g., soil nutrients) have been transformed into high 
entropy (wastes and pollution). Indeed, it is perhaps ironical but very much in the 
logic of the second law of thermodynamics that nitrate pollution and phosphate 
pollution (high entropy) are substantial forms of pollution in temperate agricul- 
ture (Whitby & Adger, 1996; Cleveland, 1994), whereas N and P are the most 
limiting nutrients for sub-Saharan agriculture. The energy cost of temperate agri- 
culture has thus increased steadily as this process of partial substitution has taken 
place and as low-entropy natural capital has been used to an always greater extent 
over time. For example, the energy cost of extracting a tonne of PR to manufac- 
ture fertilizers is increasing (Cleveland, 1994). As noted by Georgescu-Roegen 
(1971, p. 303), the price we have to pay for agricultural intensification is the 
decrease in the low-entropy natural capital of the globe. Furthermore, "substitu- 
tion within a finite stock of accessible low entropy whose irrevocable degrada- 
tion is speeded up through use cannot possibly go on for ever (Georgescu- 
Roegen, 1993, p. 92)." 

The foregoing discussion on definitions of renewable and nonrenewable 
natural capital and intergenerational externalities indicates that there are two 
basic ecological-economic questions, which should be posed concerning the use 
of natural capital. In the case of renewable N capital, the fundamental question is 
how best to invest in N in order to maintain the stock and flows needed for agri- 
cultural production in any given period. In the case of nonrenewable P capital, the 
question is different. Investment in increasing this capital (PR) is impossible over 

time scales of human interest. The only question left then is how best to liquidate 
existing inventories ofP (inventories of remaining stocks). 

The perspective adopted in this chapter for addressing these two questions 
is both that of African farmers and of society at large. The stocks of N and P con- 
sidered here are thus those in various fractions of farmers' soils as well as in min- 
eral deposits. 

Rule for Using Natural Capital in African Agriculture 

In conventional mainstream economics where no limits are put onto the 
substitution of manufactured capital or natural capital, a rule or theorem has been 
derived for the optimal use of nonrenewable resources, such as minerals. This 
rule was developed by Solow (1974, p. 41), and states that "earlier generations 
are entitled to draw down the pool [of resources] (optimally of course!) so long 
as they add (optimally of course!) to the stock of reproducible [manufactured] 
capital." Optimality is defined here by reference to the rate of interest (a measure 
of the opportunity cost of capital) and the social rate of time preferences (rate at 
which society trades off the present for the future), which are concepts that need 
not enter into this argument. In other words, optimal depletion of nonrenewable 
capital requires that the marginal benefits of depletion, to those who deplete a 
stock of natural capital, be invested in increases in manufactured capital. 

This rule has recently been modified by an ecological economist to account 
for the fact that manufactured capital cannot be a full substitute for natural capi- 
tal (since natural capital is necessary to the formation of manufactured capital). 
Daly (1994) proposed that the optimal way of liquidating inventories of nonre- 
newable natural capital is for the net gains of liquidation to be used to finance 
investments in a partial substitute, namely, renewable natural capital stocks. 

This rule can be applied to the case of P and N. Phosphorus is converted in 
situ into agricultural output and in the process is being depleted. At each step of 
this conversion process the benefits from increasing agricultural output should be 
maximized while at the same time the loss of ecological flows and ecosystem ser- 
vices associated with an in situ decrease in P pools should be minimized. This can 
be achieved when the costs of P depletion (loss of ecological amenities) are part- 
ly compensated by investments in some renewable natural capital, which will 
enhance soil quality. 

Renewable natural capital is of course not a total substitute for nonrenew- 
able natural capital. Rather, some renewable natural capital can be defined as a 
partial substitute, largely complementary to some nonrenewable natural capital, 
in the sense that it can improve the efficiency of energy flows and material recy- 
cling in agroecosystems. The point of this argument is that a decrease in the stock 
of nonrenewable natural capital (P) can, from an economic cost perspective, be 
partially compensated by an investment in the stock of some renewable capital. 
This does not imply that this renewable capital should be, in this case, an agro- 


nomic substitute for P, but simply that the inexorable depletion of nonrenewable 
P should be partially palliated by increases in soil-enhancing, renewable natural 

In this specific sense, soil organic matter (and N as part of soil organic mat- 
ter), which was defined above as renewable natural capital, can be considered as 
such a partial trade-off for P for two reasons. First, it increases the efficiency of 
cycling of N, P, and K in agroecosystems. It thus ensures that P is used more effi- 
ciently in these systems. Second, it provides energy to drive the P cycle, which 
substitutes for the petrochemical energy used to manufacture fertilizer (the clos- 
est manufactured substitute). Another way of stating this is to note that the deple- 
tion of P stocks in agriculture should have two purposes: agricultural production 
in the shorter-term and soil fertility enhancement through increased investments 
in soil organic matter in the longer term. 

Daly's rule for compensating, to the extent possible, declines in nonrenew- 
able natural capital by increases in some renewable natural capital results in the 
case of the two soil nutrients under scrutiny in this paper into the following 
requirement. The best way of liquidating P inventories, both on-farm and in 
deposits, is such that the marginal costs of P depletion (for in situ P, marginal ben- 
efits for deposits) are equal (at the margin) to investments in increases in 
organic matter on farmers' fields. This assumes that from society's perspective, 
agricultural uses of P deposits are preferable to, or at least as beneficial as indu; 
trial uses of these deposits. A discussion of the appropriateness of this assump- 
tion lies outside the scope of this chapter. In thermodynamic terms this is a s 
rior solution to the conventional replacement of P and N by fertilizers (e.g., triple 
superphosphate) because it results in increases in a low-entropy resource (soil 
organic matter). Fertilizers, by comparison, are high-entropy manufactured capi 
tal, which in addition further draw down P deposits. 

To implement the rule proposed here in Africa it will be necessary to apply 
relatively large quantities of organic material together with PR coming from 
deposits to the depleted soils, and wherever large quantities of organic material 
are not feasible, applications of mineral N fertilizer (hereafter referred to as inor- 
ganic N). Excessive quantities of organic inputs, however, can be detrimental to 
crop yields and ecosystem functioning. While organic material has relatively low 
entropy, PR applications are likely to have greater entropy and inorganic N yet 
higher entropy (see Giller et al., 1997, this publication; Palm et al., 1997, this 

These applications would serve to catalyze a number of ecological func- 
tions such as decomposition, synthesis of soil organic matter, activities of soil 
biota, N and P cycles and in essence jump-start the process of agricultural pro- 
duction in agroecosystems. There are a number of PR deposits in Africa, which 
could conceivably be used as a source for these applications (Buresh et al., 1997, 
this publication). Applications of organic material can take many generic forms, 
including incorporation of compost, crop residues, litter, green manures, and bio- 
mass transfers from agroforestry and legume intercropping systems (Palm et al., 
1997, this publication). The specifics of the methods of application and quantity 
of organic materials to be applied will vary with soil type and climate and the 
socioeconomic conditions of farmers. Likewise, the quantities of PR applications 

will be determined by soil type, climate, and socioeconomic i 
(Sanchez et al., 1997, this publication). Finally, in addition to applications of 
organic matter and PR, soil erosion control measures will be needed, particularly 
on sloping lands, to ensure that the materials applied are not transported some- 
where else. 

Who Should Pay for What? 

The essential point in this argument is that applications of PR alone, and 
(but to a lesser extent) applications of organic materials alone are insufficient to 
resolve the issue of nutrient depletion in Africa. It is indeed the combination of 
these two types of interventions, plus erosion control measures where applicable 
and other nutrient applications where necessary, which are called for. Such inter- 
ventions will replenish the N part of the natural capital of African countries and 
will use up efficiently that part of this capital (P) that is nonrenewable. In addi- 
tion to, and as a consequence of the catalytic role played by such applications 
from an ecosystem perspective, a number of positive environmental externalities 
will result from such applications. These range from increased C sequestration in 
soils and in aboveground biomass, increased biodiversity, increased watershed 
protection and quality of water supplies, decreased deforestation, and desertifica- 
tion (in some parts of Africa) to increased food security through agricultural sus- 
tainability. The benefits of increased food security would, in turn, result in a num- 
ber of additional social benefits that include in particular rural poverty alleviation 
and employment generation through the multiplier effect. These externalities will 
occur, assuming that these combinations of investments in soil organic matter and 
applications of PR can be implemented on a sufficiently large geographical scale. 

Positive environmental externalities are defined here as follows. At a given 
time, investments in soil organic matter and applications of PR on farmers' fields 
will trigger flows of nonmonetized benefits accruing to different groups in soci- 
ety and called environmental externalities. Local farmers will accrue some of 
these benefits, largely those related to increased food security at the local scale. 
(They will receive these benefits in addition to the increased income, which will 
be a direct outcome of the interventions to replenish soil fertility). National soci- 
ety will receive the benefits of decreased deforestation and desertification, of the 
protection of watershed and water supplies as well as benefits of regional and per- 
haps in some cases national food security. Furthermore, the national benefits of 
decreased rural poverty and of a more dynamic agricultural sector are potentially 
significant; with appropriate macroeconomic policies, agricultural growth could 
be one of the engines of overall economic growth in African countries (Goldman, 
1994). Global society will enjoy mainly increased C sequestration and biodiver- 
sity benefits. 

The fact that positive externalities will result from the large-scale imple- 
mentation of the proposed rule begs the question of who should pay for this 
implementation. Clearly, the costs of investment rates of PR applications are cur- 
rently beyond the reach of the vast majority of farmers in the African continent. 
Phosphate rock deposits in Togo, for example, are mined largely for export mar- 
kets, and it has been hypothesized that poverty levels in the agricultural sector 


have hampered the development of an internal demand for soil fertility amend- 
ments in most African countries (Perkins & Roemer, 1994). In any case, the exis- 
tence of externalities indicates that it would not be optimal nor effective nor equi- 
table to expect African farmers to bear the totality of these costs. Figure 10-2 
illustrates this situation. The marginal costs of these applications (MC) are com- 
pared with the marginal benefits received by individual farmers (MB,), those 
received by the national society (MB„) and those received by the global society 
(MB g ). The exact shapes of the marginal benefit functions are an empirical issue. 
Some very preliminary evidence indicates that the respective marginal benefit 
functions represented in Fig. 10-2 may be acceptable approximations. Three case 
studies of the economic and environmental costs and benefits of the use of PR in 
Madagascar, Zimbabwe, and Burkina Faso were recently undertaken under the 
auspices of the World Bank. In Zimbabwe and Burkina Faso, it would appear that 
the global benefits of such applications are indeed greater than the national and 
individual ones, at the margin, and as indicated in Fig. 10-2 (Johnsen et al, 
1996). The case of Madagascar is somewhat different in that PR deposits are 
located in islands that are classified as biodiversity hot spots on the World 
Heritage list. Exploiting these deposits would thus have significant environmen- 
tal costs: a decrease in biodiversity levels. 

For optimal levels of investment in soil nutrients (Q g in Fig. 10-2) to com- 
pensate, to a certain extent, for P depletion, policy interventions will be needed. 
In the absence of such interventions farmers will be willing to invest only up to 
Q a , assuming that they are actually able to do so, which in most cases in Africa 
is a heroic assumption. We know that high levels of rural poverty are indeed a 
constraint to any technological adoption in Africa. In 1990, for example, 30 coun- 
tries in Africa had a gross national product (GNP) per capita of less than US$ 500 
per year, while six countries had a GNP per capita of between US$ 500 and US$ 
1000, and three countries had a GNP per capita between US$ 1010 and US$ 3330 
(Tomich et al., 1995, p. 11-12). Such income levels do not leave much room for 

-ig. 10-2. Marginal benefit (MB) and marginal cost (MC) from P and N received by in 
ers (i), the national society (n), and the global society (g). The optimal level of inve 
nutrients is represented by Qi for individual farmers. (>., for the national society, a 
global society. 

any form of on-farm investments by farmers. This is why many authors consider 
that rural credit schemes are a prerequisite for any improvement in African agri- 
culture (Cole & Duesenberry, 1994). In addition to such credit schemes, howev- 
er, policy interventions are needed to implement optimal levels of soil replenish- 

An equitable principle or goal for these interventions would be to spread 
the financial burden of investments in soil nutrients across the groups in society 
that benefit from this investment. This principle of the beneficiary pays would 
require that the global society contribute financially to these investments by an 
amount equal the flows of benefits it receives, which are probably greater that 
the flows received by other groups in society, as seen above. National societies 
would likewise provide funds equivalent to die flows of benefits they receive. 
Farmers would incur the remaining costs. Such a cost-sharing principle is not 
new, and there already are some mechanisms in place for its implementation. 

At the global level, the Global Environmental Facility (GEF) is a mecha- 
nism for transferring funds from countries in the North to countries in the South, 
which undertake activities that generate global biodiversity and climate change 
benefits. GEF is thus the very means of implementing the cost-sharing principle 
proposed here between global and national societies. Another existing way to 
make such international transfers is C offsets. They are a type ofjoint imple- 
mentation in which a developed country's power utility and a developing coun- 
try's Forestry Department come to an agreement whereby the power utility 
finances reforestation efforts in the developing country in exchange for credit for 
the C sequestered by the reforestation. Such a mechanism could be extended to 
other undertakings, in addition to reforestation, and include investments in soil 
replenishments. The World Bank offers anodier possibility for implementing this 
cost-sharing principle through the granting of loans carrying no interest. 

At the national level, this cost-sharing principle can be put into practice by 
designing a number of measures aiming at offering farmers payments in return 
for adopting certain management practices. The Organization for Economic Co- 
operation and Development (OECD) countries, for example, have recently put in 
place a number of mechanisms (subsidies, compensatory payments, and special 
grants) for ensuring that their farmers adopt practices that are environmentally 
sound (OECD, 1993). Compensatory payments are associated with restrictions 
on farming practices (e.g., to reduce nitrate and pesticide leaching into ground- 
waters); these compensatory payments are based on the principle of profits fore- 
gone (OECD, 1993, p. 53). Schemes for encouraging farmers to adopt environ- 
mentally sound management practices (e.g., integrated pest management; contour 
plowing) entail payments to farmers who adopt. The payments are directly relat- 
ed to the costs to the farmers of the form of management required, including the 
opportunity costs of this management (OECD, 1993, p. 54). A third policy instru- 
ment often used in OECD countries are special grants provided to farmers who 
wish to change their methods of production, from orthodox farming methods to 
alternative or organic farming. The grants cover the period of changeover during 
which the income of the farmers is likely to fall (OECD, 1993, p. 54). Finally, a 
mechanism that is used by the majority of European countries is management 
agreements. These take the form of a legal contract between public authorities 


and farmers, in which the latter receive regular payments in return for providing 
specific environmental services (OECD, 1993, p. 58). These agreements are often 
for a fixed number of years, but some are indefinite in length. 

The USA does not really use such agreements but prefers to use cost-shar- 
ing schemes, as a variant of these agreements (Garrett & Buck, 1997). For exam- 
ple, in 1989 the U.S. Department of Agriculture initiated such a scheme for farm- 
ers adopting integrated crop management practices designed to reduce pesticide 
and fertilizer use (OECD, 1993, p. 60). Another example is that American farm- 
ers who decide to undertake tree planting to meet soil conservation objectives 
qualify for cost-sharing assistance from the long-standing Agricultural 
Conservation Program (Garrett & Buck, 1997). 

It is somewhat ironical that, in addition to the mechanisms just mentioned, 
OECD countries also have put in place various means for controlling (in the sense 
of reducing) the use of fertilizer and manure production by their farmers. These 
means consist of a set of controls and regulations that include, for example, bans 
on manure spreading, maximum ceilings set on livestock densities, and ceilings 
on applications of fertilizers (OECD, 1993). The nitrate and phosphate pollution 
associated with agricultural production in Europe and North America is of course 
a consequence of agricultural subsidies in these countries, which are biasing P 
extraction and use towards farms in the North. 

The point of these examples is to demonstrate that the cost-sharing princi- 
ple advocated here for sub-Saharan countries has in fact been implemented by a 
majority of countries in the North for a number of years. Different objectives 
were sought by different countries, but the fact remains that farmers in these 
countries have been receiving various sorts of payments (and new ones are con- 
stantly being developed, as agricultural research indicates that different practices 
will best address various environmental problems) as incentives for adopting 
environmentally sound practices. It also should be remembered that such pay- 
ments are taking place in countries where GNP per capita is of the order of US$ 
20 000 per year. Given this international context, arguing that farmers in Africa 
should not receive subsidies, which would interfere with structural adjustment 
programs, as has been argued by some orthodox economists, rather appears to be 
a case of "do as I say, but not as I do. 

The specific mechanisms that can be put in place in African countries to 
implement the soil replenishment strategy proposed in this chapter will vary from 
country to country, depending upon existing institutions and economic, agricul- 
tural, and environmental policies. For the purpose of this chapter it is sufficient 
to note that feasible mechanisms have been put in place both globally and in var- 
ious countries that can now be used for putting into application the cost-sharing 
principle proposed for this impler 


A substantial amount of research is needed before the rule proposed here 
for investing in the renewable natural capital nutrient of African countries can 
become operational. There are six principal research themes that warrant further 

Identification of the Soils that Should Receive Nutrient Investments 

The first type of data needed concerns the characteristics and the geo- 
graphical location of the soils that require investments in nutrient capital. The 
combination of measures advocated here (applications of PR and organic materi- 
als, soil erosion control measures) needs to be targeted to specific soils. Those are 
presumably P depleted, including P-fixing soils with depleted soil organic matter. 
The problem is that not sufficient information has currently been compiled for the 
identification of such target soils to be possible in the majority of African coun- 
tries. It is; however, essential that the geographical location of these soils be 
known before soil replenishment can actually be implemented. Research involv- 
ing soil scientists and geographers with skills in geographic information systems 
and remote sensing is therefore needed to generate this type of information. 

Most Effective Combination of Applications 

Secondly, it is necessary to determine the most effective combinations of 
applications of PR, organic materials, inorganic N, and soil erosion measures for 
the different types of soils identified as targets for nutrient replenishment. 
Effectiveness is defined here in terms of efficiency of energy flows and material 
recycling within agroecosystems, in keeping with the ecological-economic argu- 
ment developed in this chapter concerning the trade-offs between nonrenewable 
and renewable natural capital. The advocated combination of applications will of 
course vary significantly in terms of actual quantities applied, frequency of appli- 
cations, type of organic inputs to be used, and overall management regimes with 
differences in soil types and climate (Giller et al, 1997, this publication; Palm et 
al, 1997, this publication). Soil scientists, soil biologists, systems agronomists, 
and ecologists will have to undertake this type of research. 

Flows of Ecological Amenities 

Once the appropriate combinations of applications are identified for the 
major target soils in a country, it will be necessary to quantify the flows of eco- 
logical amenities (e.g., increased C sequestration, biodiversity, decreased defor- 
estation, desertification) that will be associated with each different combination 
over a period of about 7 to 10 yr. The same disciplines as above will contribute 
to such research. Valuations of these flows of amenities will then be needed, to 
determine the relative share of local versus national versus global flows. Only 
approximate values will be required for this purpose. A number of methods are 
available and have been used to assess the monetary worth of amenities such as 
C sequestration (see Pethig, 1994, for details). Ecological economists will be 
involved in these valuations. 

Economic Reserves 

A fourth area for further research concerns deposits of PR in Africa. An 
empirical issue is whether there are enough reserves of PR deposits of a sufficient 


quality in the continent. A related issue is how long these reserves will last. In 
other words, is the investment scenario proposed in this chapter and by other 
authors in the symposium a long-term solution, or is it rather a short-term solu- 
tion because reserves that are economic and profitable to exploit are not signifi- 
cant? And finally, what are the costs of exploiting these reserves and of trans- 
porting PR to the different target areas and soils in a given country? Geologists, 
mining engineers, and economists will provide such information. 

Comparison of Processed Phosphorus Fertilizer 
and Natural Capital Investments 

On the basis of all of the above information, it will be possible to compare 
the investment in natural capital advocated here with applications of processed P 
fertilizers (such as triple superphosphate or whatever happens to be the next best 
alternative) in terms of their respective energetic efficiency, flows of ecological 
amenities, and flows of costs and benefits at various spatial scales, including food 
security benefits and economic growth benefits at the national scale. Economists 
and natural scientists will contribute to such a comparison. This comparison will 
be needed for determining which groups in society are the winners and the losers 
when natural capital investments are implemented and when processed P fertiliz- 
er is applied. This in turn is essential information for devising an effective imple- 
mentation scheme in each country. 

Operational Cost-Sharing Mechanism 

On the basis of the results yielded by the above research themes, an opera- 
tional cost-sharing mechanism will have to be developed by each country. This 
mechanism will be shaped by the specific policy and institutional environment of 
the country concerned. One of the issues it will need to address is that of the 
socioeconomic characteristics of the farms that should be targeted for soil replen- 
ishment efforts. As PR reserves are highly unlikely to be sufficient for applica- 
tions of combinations of PR and organic material to be feasible on all target soils, 
and as the financial costs of undertaking these applications on all target soils 
(some of which will presumably be in very remote areas with poor transport facil- 
ities) are likely to be prohibitive when compared with the funds that can be raised 
through international and national transfer mechanisms, some priorities will have 
to be set. Economic theory would indicate that the farms that should be targeted 
are those for which returns to farmers, the national, and the global society are the 
highest (principle of the biggest bang for a given investment). 

Trade-offs are likely to exist between the interests of different groups or 
categories of farmers. For example, applications of PR and organic materials to 
the farms of the poorest farmers could, under certain circumstances, bring about 
the highest flows of environmental amenities to the global society but relatively 
low benefits to the national society in terms of economic growth and food secu- 
rity. Similarly, applications to the farms of the better off farmers could generate 
high national economic benefits but relatively low global benefits. In such cases, 
value judgments on the part of national governments will be needed to sort out 

these trade-offs. Ethical considerations concerning the welfare of the poorest of 
the poor versus the welfare of the better off farmers also will play a role in estab- 
lishing such trade-offs. Of course, the use of international transfer mechanisms in 
implementing the investment scheme advocated here should also help to sort out 
these trade-offs by providing the national government with a direct incentive to 
take global benefits into consideration when it identifies target farmers. Policy 
scientists and government authorities representing the different groups of interest 
involved in soil replenishment in Africa would have to be involved in the devel- 
opment of such an operational cost-sharing mechanism. 


Based upon concepts from ecological economics and thermodynamics, it 
can be argued that P is nonrenewable natural capital and N is renewable natural 
capital. The basic economic questions are then how best to liquidate nonrenew- 
able P inventories and how best to invest in renewable N. A theorem of main- 
stream economics concerning the optimal use of exhaustible resources, modified 
by Daly (1994) to account for the nonsubstitutability of manufactured capital for 
natural capital, was further modified and applied to the case of P and N. This rule 
stipulates that a socially optimal way of liquidating P inventories and investing in 
N is to invest in soil organic matter — renewable natural capital with low entropy 
— by an amount equal (at the margin) to the costs of depleting P. 

The implementation of this rule necessitates applications of PR and organ- 
ic materials, in conjunction with appropriate accompanying measures such as soil 
conservation and application of inorganic N fertilizer. Such interventions are 
beyond the financial reach of most farmers in Africa. However, it can be argued 
that given the positive environmental and social externalities associated with 
these interventions, an equitable mechanism would be to share the costs of their 
implementation among the different groups in society that benefit. 

A quote from Holling (1994, p. 72) appropriately encapsulates the com- 
plexity of the issues discussed in this chapter. "Sustaining the biosphere [and nat- 
ural capital] is not an ecological problem, or a social problem, or an economic 
problem. It is an integrated combination of all three." 


The comments and inputs of Mike Swift, Roland Buresh, Keith Shepherd, 
Frank Place, and Meredith Soule on an earlier draft of this chapter are gratefully 


Bekunda, M.A., A. Bationo, and H. Ssali. 1997. Soil fertility management in Africa: A review of 
leered i rch I I i ( ' 1 I L i li et at. (ed Repleni liii il 1 til ■ in 

Africa. SSSA Spec. Publ. 51. SSSA, Madison, WI (this publication). 

Bruno. A. 1930. La restauration tie la fertilite lies terres tropicales. Bull. Aerie. Contio Beiize. 


Buresh, R.J., P.C. Smithson, and D.T. Heliums. 1997. Building soil phosphorus capital in Africa, p. 

MM In R.J. Buresh et al i Repleni hin il ft tilii i Africa ^ Spec. Publ. 51. 

SSSA, Madison, WI (this publication). 
C'l land ! I > ' 1' tin il i i in ii i i n u I i il i multure: Examples 

rom hull i lid i i ' nil i in p 1 >- 1 1 ) >/ > lansson et al. (ed.l In ting ii n itui I 

. ipil il Island Pi V i lington I > 
Cole, D.C., and J S Dueseuberr} i" S Financi In l> mnent, p. 305-336. In D.L. Lindaner and M. 

Roemer (ed.) Asia and Africa, legacies and opportunities in development. Inst, for 

Contemporan Stud. Press. San Francisco. 
Daly. Ill I ' >i ii i il i i Ml lopinent I n i n i tin il capital, p. 22-37. 

/// A. .lansson et al. led ! Investing in natural capital. Island Press. Washington. DC. 
Garrett, H.E., and L. Buck. 1997. Agroforestry practice and policy in the L'nited States of America. 

For. Ecol. Manage. 91:5-15. 
Georgescu-Roegen, N. 1971. The entropy law and tile economic process. Harvard Univ. Press, 

Cambridge, MA. 
Georgescu-Roegen, N. 1993. Selections from "Energy and economic myths, 1975." p. 89-112. In 

H.E. Dal) and K.N. Townsend led.) Valuing the earth, ecology, economies and ethics. MI') 

Press, Cambridge, MA. 
Giller, K.E., G. Cadisch, C. Ehaliotis, E. Adams, W.D. Sakala, and P.L. Mafongoya. 1997. Building 

soil n h i u ] n I in ii i III ' ' I l in '\ i ii i i i j I iii Inn i I i ml 

itj in Africa. SSSA Spec. Publ. 51. SSSA, Madison, WI (this publication). 
Goldman, R.H. 1994. Patterns and sources of food systems development, p. 195-232. In D.L. 

Lindaner and M. Roemer (ed.) Asia and Africa, legacies and opportunities in development. 

Inst, for Contemporan Stud. Press. San Francisco. 
Holling, C.S. 1994. New science and new investments for a sustainable biosphere, p. 57-73. In D.L. 

Lhidan i (l i I n i I i i I i m i hi i i i i in il 

Inst, for Contemporary Stud Press. San Francisco. 
lohn n I II I 1 i, ml I Mukurunibii ' ukum ml I Ru il I ' \ Phosphate rock ini- 
tiative countrx case stud\ for Zimbabwe. Ctr. for Int. Environ, and Dev. Stud.. Agricultural 

Univ. of Norway, As. 
i' D. 19 i ultui 1 nd n u mi in iM h in' i a oi Recent] i md new directions. 

Organ, for Econ. Coop, and Dev., Paris. 
' lui C.A.. R.l.l i i n I l ndwa. 19 < ii I in i u i i a nd inoiganic nutrient 

sources toi soil r it hi maim i n nil i pi Inn in j I I" In R.J. Buresh et al. (ed.) 

Replenishing soil fertility in Africa. SSSA Spec. Publ. 51. SSSA, Madison, WI (this publica- 

Perkins, D.H., and M. Roemer. 1994. Differing endowments and historical legacies, p. 25-58. In D.L. 

Lindaner and M. Roemer (ed.) Asia and Africa, legacies and opportunities in development. 

Inst, for Contemporary stud. Press. San Francisco. 
Pethig, R. (ed.). 1994. Valuing the environment: Methodological and measurement issues. Kluwer 

Academic Publ.. Dordrecht, die Netherlands. 
Rhodes, E., A. Bationo, E.M.A. Smaling, and C. Visker, 1996. Nutrient stocks and flows in West 
African soils. | /I I! e et al I.) Restorin naintainii prodii 

tivity of West African soil \ i i i i I i I pment Misc. Fert. Stud. 14. Int. Fert. 

Dev. Ctr.-Africa, Lome, Togo. 
Sanchez, PA., A-M.N. Izac, I. Valencia, and C. Pieri. 1996. Soil fertility replenishment in Africa: A 
i i n i i mi i ; i i I ii renter iin from i roll ii i 

ments in Atnca ika >in cm Mexico City. 

Sanchez, P.A., K.D. Shepherd, M.J. Soule, F.M. Place, R.J. Buresh, A.-M.N. Izac, A.U. Mokwunye, 

F.R. Kwesiga, C.G. Ndiritu, and PL. Woomer. 1997. Soil fertility i pleni Inn in in \fri i >n 
ivestment in natural 1 1 ii I I--I //; R.J. Bu i et al. (ed.l Replenishin il fei 

tility in Africa. SSSA Spec. Publ. 51. SSSA, Madison, WI (this publication). 
Smaling, E.M.A., S.M. Nandwa, and B.H In n Miu i M i il i ike p 47-61. In R.J. 

Buresh etal. (ed.i Replenishing soil fertility in Africa. SSSA Spec. Publ. 51. SSSA, Madison, 

WI (this publication). 
Solow, R.M. 1974. Intergenerational equity and exhaustible resources. Rev. Econ. Stud. 41:139-152. 
Stoorvogel, J.J., E.M.A. Smaling. aim B.H. Janssen. 199 I Cah Mating soil nutrient balances in Africa 

i Mi in i [.Supra nati mal scale. Fert. Res. 35:227-235. 
Tomicli. T.P.. P. Kilbv. and B.F. Johnston. 1995. Transforming agrarian economies. Cornell Univ. 

Press, Ithaca, NY. 
Whitby, M., and W.N. Ad°er. 1996. Natural and reprodui i hie capital ami the sustainability of land use 

in the UK. J. Agric. Econ. 47:50-65.